AI/rocm-systems
2
0
Garpu 0
Kode Masalah Tarik Permintaan Actions Packages Projects Rilis Wiki Kegiatan

SWDEV-490062 - Update documentation

Telusuri Sumber 
Change-Id: Ib5297fdda2e05795b3b20436cc1de962e310b08b


[ROCm/hip commit: 3d60bd3a64]
This commit is contained in:
Istvan Kiss
2024-10-21 16:50:09 +02:00
committed by Istvan Kiss
orang tua 862e713a5c
melakukan 122cecc59d
108 mengubah file dengan 14036 tambahan dan 2987 penghapusan
Download Patch File Download Diff File Expand all files Collapse all files
projects/hip/.github/dependabot.yml
Vendored
+1
Melihat File
@@ -14,5 +14,6 @@ updates:
- "documentation"
- "dependencies"
- "ci:docs-only"
target-branch: "docs/develop"
reviewers:
- "samjwu"
projects/hip/.readthedocs.yaml
+10 -10
Melihat File
@@ -6,28 +6,28 @@ version: 2
sphinx:
configuration: docs/conf.py
formats: [htmlzip, pdf, epub]
formats: []
python:
install:
- requirements: docs/sphinx/requirements.txt
conda:
environment: docs/environment.yml # needed until ubuntu ships doxygen >= 1.9.8
build:
os: ubuntu-22.04
tools:
python: "3.10"
python: "mambaforge-22.9" # needed until ubuntu ships doxygen >= 1.9.8
apt_packages:
- "doxygen"
- "gfortran" # For pre-processing fortran sources
- "graphviz" # For dot graphs in doxygen
jobs:
post_checkout:
- if [ -d ../llvm-project ]; then rmdir ../llvm-project; fi
- if [ -d ../clr ]; then rmdir ../clr; fi
- if [ -d ../ROCR-Runtime ]; then rmdir ../ROCR-Runtime; fi
- git clone --depth=1 --single-branch --branch rocdoc-195 https://github.com/StreamHPC/llvm-project.git ../llvm-project
- git clone --depth=1 --single-branch --branch develop https://github.com/ROCm/clr.git ../clr
- git clone --depth=1 --single-branch --branch master https://github.com/ROCm/ROCR-Runtime.git ../ROCR-Runtime
- if [ -d ../clr ]; then rm -rf ../clr; fi
- if [ -d ../ROCR-Runtime ]; then rm -rf ../ROCR-Runtime; fi
- git clone --depth=1 --single-branch --branch docs/develop https://github.com/ROCm/clr.git ../clr
- git clone --depth=1 --single-branch --branch master https://github.com/ROCm/ROCR-Runtime.git ../ROCR-Runtime
post_build:
- rm -rf ../clr
- rm -rf ../llvm-project
- rm -rf ../ROCR-Runtime
projects/hip/.wordlist.txt
+59 -2
Melihat File
@@ -1,8 +1,11 @@
.hip_fatbin
ALU
ALUs
AmgX
APU
APUs
AQL
AXPY
Asynchrony
backtrace
Bitcode
@@ -12,40 +15,65 @@ builtins
Builtins
CAS
clr
coroutines
Ctx
cuBLASLt
cuCtx
CUDA's
cuDNN
cuModule
dataflow
deallocate
decompositions
denormal
Dereferencing
dll
DirectX
EIGEN
EIGEN's
enqueue
enqueues
entrypoint
entrypoints
enum
enums
embeded
extern
fatbin
fatbinary
foundationally
frontends
fnuz
FNUZ
fp
gedit
GPGPU
GWS
hardcoded
HC
hcBLAS
HIP-Clang
HIP's
hipcc
hipCtx
hipexamine
hipified
hipModule
hipModuleLaunchKernel
hipother
HIPRTC
hcBLAS
icc
IILE
iGPU
inplace
Interoperation
interoperate
Interprocess
interprocess
Intrinsics
intrinsics
IPC
IPs
isa
Lapack
latencies
@@ -59,34 +87,63 @@ ltrace
makefile
Malloc
malloc
memset
multicore
multigrid
multithreading
multitenant
MALU
NaN
NCCL
NDRange
nonnegative
NOP
Numa
Nsight
ocp
overindex
overindexing
oversubscription
pixelated
pragmas
preallocated
preconditioners
predefining
prefetched
preprocessor
PTX
PyHIP
queryable
prefetching
quad
representable
RMW
ROCm's
rocTX
roundtrip
RTC
RTTI
rvalue
SAXPY
scalarizing
sceneries
shaders
SIMT
SOMA
SPMV
structs
SYCL
syntaxes
texel
texels
tradeoffs
templated
toolkits
typedefs
unintuitive
UMM
unmap
upscaled
variadic
WinGDB
zzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz
zzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz
projects/hip/README.md
+2 -2
Melihat File
@@ -37,7 +37,7 @@ HIP releases are typically naming convention for each ROCM release to help diffe
* [Installation](docs/install/install.rst)
* [HIP FAQ](docs/how-to/faq.md)
* [HIP Kernel Language](docs/reference/kernel_language.rst)
* [HIP C++ Language Extensions](docs/reference/cpp_language_extensions.rst)
* [HIP Porting Guide](docs/how-to/hip_porting_guide.md)
* [HIP Porting Driver Guide](docs/how-to/hip_porting_driver_api.md)
* [HIP Programming Guide](docs/how-to/programming_manual.md)
@@ -88,7 +88,7 @@ hipMemcpy(C_h, C_d, Nbytes, hipMemcpyDeviceToHost);
The HIP kernel language defines builtins for determining grid and block coordinates, math functions, short vectors,
atomics, and timer functions.
It also specifies additional defines and keywords for function types, address spaces, and optimization controls (See the [HIP Kernel Language](docs/reference/kernel_language.rst) for a full description).
It also specifies additional defines and keywords for function types, address spaces, and optimization controls (See the [HIP C++ Language Extensions](docs/reference/cpp_language_extensions.rst) for a full description).
Here's an example of defining a simple 'vector_square' kernel.
```cpp
projects/hip/docs/conf.py
+9 -1
Melihat File
@@ -43,4 +43,12 @@ extensions += ["sphinxcontrib.doxylink"]
cpp_id_attributes = ["__global__", "__device__", "__host__", "__forceinline__", "static"]
cpp_paren_attributes = ["__declspec"]
suppress_warnings = ["etoc.toctree"]
suppress_warnings = ["etoc.toctree"]
numfig = False
exclude_patterns = [
"doxygen/mainpage.md",
"understand/glossary.md"
]
projects/hip/docs/data/how-to/cooperative_groups/thread_hierarchy_coop_bottom.drawio
+904
Melihat File
@@ -0,0 +1,904 @@
<mxfile host="65bd71144e">
<diagram id="zBbb_w2fufU70cdOGtND" name="1 oldal">
<mxGraphModel dx="757" dy="937" grid="1" gridSize="10" guides="1" tooltips="1" connect="1" arrows="1" fold="1" page="1" pageScale="1" pageWidth="1200" pageHeight="1600" background="none" math="0" shadow="0">
<root>
<mxCell id="0"/>
<mxCell id="1" parent="0"/>
<mxCell id="5818" value="" style="rounded=0;fillColor=#5E5B61;fontColor=#FFFFFF;strokeColor=none;spacing=0;" vertex="1" parent="1">
<mxGeometry x="20" y="90" width="640" height="450" as="geometry"/>
</mxCell>
<mxCell id="5400" value="" style="rounded=0;fillColor=#C23555;fontColor=#FFFFFF;strokeColor=none;spacing=0;" vertex="1" parent="1">
<mxGeometry x="30" y="120" width="620" height="200" as="geometry"/>
</mxCell>
<mxCell id="5401" value="Block" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="310" y="120" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="5402" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#A20025;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="360" y="160" width="280" height="140" as="geometry"/>
</mxCell>
<mxCell id="5403" value="&lt;font face=&quot;Klavika&quot; style=&quot;font-size: 17px;&quot;&gt;Thread-block tile&lt;/font&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="420" y="160" width="160" height="30" as="geometry"/>
</mxCell>
<mxCell id="5404" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="390" y="210" width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="5405" value="Warp" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="405" y="210" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="5406" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="399.8600000000001" y="236.5" as="sourcePoint"/>
<mxPoint x="399.8600000000001" y="283.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5407" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="409.8600000000001" y="236.5" as="sourcePoint"/>
<mxPoint x="409.8600000000001" y="283.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5408" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="419.8600000000001" y="236.5" as="sourcePoint"/>
<mxPoint x="419.8600000000001" y="283.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5409" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="429.8600000000001" y="236.5" as="sourcePoint"/>
<mxPoint x="429.8600000000001" y="283.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5410" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="439.8600000000001" y="236.5" as="sourcePoint"/>
<mxPoint x="439.8600000000001" y="283.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5411" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="449.8600000000001" y="236.5" as="sourcePoint"/>
<mxPoint x="449.8600000000001" y="283.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5412" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="459.8600000000001" y="236.5" as="sourcePoint"/>
<mxPoint x="459.8600000000001" y="283.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5413" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="469.8600000000001" y="236.5" as="sourcePoint"/>
<mxPoint x="469.8600000000001" y="283.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5414" value="" style="group;fontColor=#FFFFFF;" vertex="1" connectable="0" parent="1">
<mxGeometry x="380" y="200" width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="5415" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#523E43;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="5414">
<mxGeometry width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="5416" value="Warp" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontColor=#FFFFFF;" vertex="1" parent="5414">
<mxGeometry x="15" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="5417" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5414">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="9.86" y="26.5" as="sourcePoint"/>
<mxPoint x="9.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5418" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5414">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="19.860000000000113" y="26.5" as="sourcePoint"/>
<mxPoint x="19.860000000000113" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5419" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5414">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="29.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="29.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5420" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5414">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="39.86000000000011" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="39.86000000000011" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5421" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5414">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="49.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="49.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5422" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5414">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="59.86000000000011" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="59.86000000000011" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5423" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5414">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="69.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="69.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5424" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5414">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="79.86000000000013" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="79.86000000000013" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5425" value="" style="group;fontColor=#FFFFFF;" vertex="1" connectable="0" parent="1">
<mxGeometry x="370" y="190" width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="5426" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#585556;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="5425">
<mxGeometry width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="5427" value="Warp" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="5425">
<mxGeometry x="15" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="5428" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5425">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="9.86" y="26.5" as="sourcePoint"/>
<mxPoint x="9.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5429" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5425">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="19.860000000000113" y="26.5" as="sourcePoint"/>
<mxPoint x="19.860000000000113" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5430" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5425">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="29.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="29.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5431" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5425">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="39.86000000000011" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="39.86000000000011" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5432" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5425">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="49.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="49.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5433" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5425">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="59.86000000000011" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="59.86000000000011" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5434" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5425">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="69.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="69.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5435" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5425">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="79.86000000000013" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="79.86000000000013" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5436" value="" style="group;fontColor=#FFFFFF;" vertex="1" connectable="0" parent="1">
<mxGeometry x="540" y="210" width="100" height="90" as="geometry"/>
</mxCell>
<mxCell id="5437" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="5436">
<mxGeometry width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="5438" value="Warp" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontColor=#FFFFFF;" vertex="1" parent="5436">
<mxGeometry x="15" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="5439" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5436">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="9.86" y="26.5" as="sourcePoint"/>
<mxPoint x="9.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5440" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5436">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="19.860000000000113" y="26.5" as="sourcePoint"/>
<mxPoint x="19.860000000000113" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5441" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5436">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="29.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="29.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5442" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5436">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="39.86000000000011" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="39.86000000000011" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5443" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5436">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="49.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="49.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5444" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5436">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="59.86000000000011" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="59.86000000000011" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5445" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5436">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="69.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="69.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5446" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5436">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="79.86000000000013" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="79.86000000000013" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5447" value="" style="group;fontColor=#FFFFFF;" vertex="1" connectable="0" parent="1">
<mxGeometry x="530" y="200" width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="5448" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#523E43;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="5447">
<mxGeometry width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="5449" value="Warp" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontColor=#FFFFFF;" vertex="1" parent="5447">
<mxGeometry x="15" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="5450" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5447">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="9.86" y="26.5" as="sourcePoint"/>
<mxPoint x="9.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5451" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5447">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="19.860000000000113" y="26.5" as="sourcePoint"/>
<mxPoint x="19.860000000000113" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5452" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5447">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="29.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="29.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5453" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5447">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="39.86000000000011" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="39.86000000000011" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5454" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5447">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="49.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="49.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5455" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5447">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="59.86000000000011" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="59.86000000000011" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5456" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5447">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="69.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="69.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5457" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5447">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="79.86000000000013" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="79.86000000000013" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5458" value="" style="group;fontColor=#FFFFFF;" vertex="1" connectable="0" parent="1">
<mxGeometry x="520" y="190" width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="5459" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#585556;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="5458">
<mxGeometry width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="5460" value="Warp" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="5458">
<mxGeometry x="15" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="5461" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5458">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="9.86" y="26.5" as="sourcePoint"/>
<mxPoint x="9.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5462" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5458">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="19.860000000000113" y="26.5" as="sourcePoint"/>
<mxPoint x="19.860000000000113" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5463" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5458">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="29.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="29.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5464" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5458">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="39.86000000000011" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="39.86000000000011" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5465" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5458">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="49.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="49.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5466" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5458">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="59.86000000000011" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="59.86000000000011" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5467" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5458">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="69.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="69.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5468" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5458">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="79.86000000000013" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="79.86000000000013" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5469" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#A20025;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="40" y="160" width="280" height="140" as="geometry"/>
</mxCell>
<mxCell id="5470" value="&lt;font face=&quot;Klavika&quot; style=&quot;font-size: 17px;&quot;&gt;Thread-block tile&lt;/font&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="100" y="160" width="160" height="30" as="geometry"/>
</mxCell>
<mxCell id="5471" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="70" y="210" width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="5472" value="Warp" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="85" y="210" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="5473" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="79.86000000000001" y="236.5" as="sourcePoint"/>
<mxPoint x="79.86000000000001" y="283.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5474" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="89.86000000000013" y="236.5" as="sourcePoint"/>
<mxPoint x="89.86000000000013" y="283.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5475" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="99.86000000000001" y="236.5" as="sourcePoint"/>
<mxPoint x="99.86000000000001" y="283.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5476" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="109.86000000000013" y="236.5" as="sourcePoint"/>
<mxPoint x="109.86000000000013" y="283.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5477" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="119.86000000000001" y="236.5" as="sourcePoint"/>
<mxPoint x="119.86000000000001" y="283.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5478" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="129.86000000000013" y="236.5" as="sourcePoint"/>
<mxPoint x="129.86000000000013" y="283.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5479" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="139.86" y="236.5" as="sourcePoint"/>
<mxPoint x="139.86" y="283.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5480" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="149.86000000000013" y="236.5" as="sourcePoint"/>
<mxPoint x="149.86000000000013" y="283.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5481" value="" style="group;fontColor=#FFFFFF;" vertex="1" connectable="0" parent="1">
<mxGeometry x="60" y="200" width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="5482" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#523E43;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="5481">
<mxGeometry width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="5483" value="Warp" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontColor=#FFFFFF;" vertex="1" parent="5481">
<mxGeometry x="15" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="5484" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5481">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="9.86" y="26.5" as="sourcePoint"/>
<mxPoint x="9.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5485" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5481">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="19.860000000000113" y="26.5" as="sourcePoint"/>
<mxPoint x="19.860000000000113" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5486" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5481">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="29.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="29.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5487" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5481">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="39.86000000000011" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="39.86000000000011" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5488" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5481">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="49.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="49.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5489" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5481">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="59.86000000000011" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="59.86000000000011" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5490" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5481">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="69.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="69.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5491" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5481">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="79.86000000000013" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="79.86000000000013" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5492" value="" style="group;fontColor=#FFFFFF;" vertex="1" connectable="0" parent="1">
<mxGeometry x="50" y="190" width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="5493" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#585556;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="5492">
<mxGeometry width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="5494" value="Warp" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="5492">
<mxGeometry x="15" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="5495" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5492">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="9.86" y="26.5" as="sourcePoint"/>
<mxPoint x="9.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5496" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5492">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="19.860000000000113" y="26.5" as="sourcePoint"/>
<mxPoint x="19.860000000000113" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5497" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5492">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="29.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="29.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5498" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5492">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="39.86000000000011" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="39.86000000000011" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5499" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5492">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="49.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="49.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5500" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5492">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="59.86000000000011" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="59.86000000000011" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5501" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5492">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="69.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="69.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5502" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5492">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="79.86000000000013" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="79.86000000000013" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5503" value="" style="group;fontColor=#FFFFFF;" vertex="1" connectable="0" parent="1">
<mxGeometry x="220" y="210" width="100" height="90" as="geometry"/>
</mxCell>
<mxCell id="5504" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="5503">
<mxGeometry width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="5505" value="Warp" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontColor=#FFFFFF;" vertex="1" parent="5503">
<mxGeometry x="15" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="5506" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5503">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="9.86" y="26.5" as="sourcePoint"/>
<mxPoint x="9.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5507" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5503">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="19.860000000000113" y="26.5" as="sourcePoint"/>
<mxPoint x="19.860000000000113" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5508" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5503">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="29.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="29.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5509" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5503">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="39.86000000000011" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="39.86000000000011" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5510" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5503">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="49.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="49.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5511" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5503">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="59.86000000000011" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="59.86000000000011" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5512" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5503">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="69.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="69.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5513" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5503">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="79.86000000000013" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="79.86000000000013" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5514" value="" style="group;fontColor=#FFFFFF;" vertex="1" connectable="0" parent="1">
<mxGeometry x="210" y="200" width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="5515" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#523E43;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="5514">
<mxGeometry width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="5516" value="Warp" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontColor=#FFFFFF;" vertex="1" parent="5514">
<mxGeometry x="15" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="5517" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5514">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="9.86" y="26.5" as="sourcePoint"/>
<mxPoint x="9.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5518" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5514">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="19.860000000000113" y="26.5" as="sourcePoint"/>
<mxPoint x="19.860000000000113" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5519" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5514">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="29.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="29.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5520" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5514">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="39.86000000000011" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="39.86000000000011" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5521" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5514">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="49.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="49.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5522" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5514">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="59.86000000000011" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="59.86000000000011" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5523" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5514">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="69.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="69.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5524" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5514">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="79.86000000000013" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="79.86000000000013" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5525" value="" style="group;fontColor=#FFFFFF;" vertex="1" connectable="0" parent="1">
<mxGeometry x="200" y="190" width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="5526" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#585556;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="5525">
<mxGeometry width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="5527" value="Warp" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="5525">
<mxGeometry x="15" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="5528" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5525">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="9.86" y="26.5" as="sourcePoint"/>
<mxPoint x="9.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5529" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5525">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="19.860000000000113" y="26.5" as="sourcePoint"/>
<mxPoint x="19.860000000000113" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5530" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5525">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="29.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="29.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5531" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5525">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="39.86000000000011" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="39.86000000000011" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5532" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5525">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="49.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="49.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5533" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5525">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="59.86000000000011" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="59.86000000000011" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5534" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5525">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="69.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="69.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5535" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5525">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="79.86000000000013" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="79.86000000000013" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5536" value="" style="rounded=0;fillColor=#C23555;fontColor=#FFFFFF;strokeColor=none;spacing=0;" vertex="1" parent="1">
<mxGeometry x="30" y="340" width="620" height="180" as="geometry"/>
</mxCell>
<mxCell id="5537" value="Block" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="310" y="340" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="5538" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#A20025;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="370" y="380" width="260" height="120" as="geometry"/>
</mxCell>
<mxCell id="5539" value="&lt;font face=&quot;Klavika&quot;&gt;Coalesced group&lt;/font&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="420" y="380" width="160" height="30" as="geometry"/>
</mxCell>
<mxCell id="5541" value="Warp" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="405" y="430" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="5561" value="" style="group;fontColor=#FFFFFF;" vertex="1" connectable="0" parent="1">
<mxGeometry x="380" y="410" width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="5562" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#585556;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="5561">
<mxGeometry width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="5563" value="Warp" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="5561">
<mxGeometry x="15" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="5564" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5561">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="9.86" y="26.5" as="sourcePoint"/>
<mxPoint x="9.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5565" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5561">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="19.860000000000113" y="26.5" as="sourcePoint"/>
<mxPoint x="19.860000000000113" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5567" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5561">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="39.86000000000011" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="39.86000000000011" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5568" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5561">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="49.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="49.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5569" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5561">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="59.86000000000011" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="59.86000000000011" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5570" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5561">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="69.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="69.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5594" value="" style="group;fontColor=#FFFFFF;" vertex="1" connectable="0" parent="1">
<mxGeometry x="530" y="410" width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="5595" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#585556;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="5594">
<mxGeometry width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="5596" value="Warp" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="5594">
<mxGeometry x="15" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="5597" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5594">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="9.86" y="26.5" as="sourcePoint"/>
<mxPoint x="9.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5598" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5594">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="19.860000000000113" y="26.5" as="sourcePoint"/>
<mxPoint x="19.860000000000113" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5599" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5594">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="29.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="29.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5600" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5594">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="39.86000000000011" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="39.86000000000011" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5602" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5594">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="59.86000000000011" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="59.86000000000011" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5603" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5594">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="69.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="69.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5604" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5594">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="79.86000000000013" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="79.86000000000013" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5605" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#A20025;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="40" y="380" width="260" height="120" as="geometry"/>
</mxCell>
<mxCell id="5606" value="&lt;font face=&quot;Klavika&quot;&gt;Coalesced group&lt;/font&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="100" y="380" width="160" height="30" as="geometry"/>
</mxCell>
<mxCell id="5608" value="Warp" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="85" y="430" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="5628" value="" style="group;fontColor=#FFFFFF;" vertex="1" connectable="0" parent="1">
<mxGeometry x="50" y="410" width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="5629" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#585556;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="5628">
<mxGeometry width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="5630" value="Warp" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="5628">
<mxGeometry x="15" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="5631" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5628">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="9.86" y="26.5" as="sourcePoint"/>
<mxPoint x="9.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5633" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5628">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="29.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="29.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5634" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5628">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="39.86000000000011" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="39.86000000000011" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5635" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5628">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="49.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="49.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5637" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5628">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="69.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="69.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5661" value="" style="group;fontColor=#FFFFFF;" vertex="1" connectable="0" parent="1">
<mxGeometry x="200" y="410" width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="5662" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#585556;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="5661">
<mxGeometry width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="5663" value="Warp" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="5661">
<mxGeometry x="15" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="5664" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5661">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="9.86" y="26.5" as="sourcePoint"/>
<mxPoint x="9.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5666" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5661">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="29.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="29.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5667" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5661">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="39.86000000000011" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="39.86000000000011" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5669" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5661">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="59.86000000000011" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="59.86000000000011" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5819" value="Grid" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="310" y="90" width="60" height="30" as="geometry"/>
</mxCell>
</root>
</mxGraphModel>
</diagram>
</mxfile>
projects/hip/docs/data/how-to/cooperative_groups/thread_hierarchy_coop_bottom.svg
+1
Melihat File
File diff suppressed because one or more lines are too long
Side by Side

After

Width:  |  Height:  |  Ukuran: 64 KiB

projects/hip/docs/data/understand/programming_model_reference/thread_hierarchy_coop.drawio → projects/hip/docs/data/how-to/cooperative_groups/thread_hierarchy_coop_top.drawio
+45 -45
Melihat File
@@ -1,6 +1,6 @@
<mxfile host="65bd71144e">
<diagram id="zBbb_w2fufU70cdOGtND" name="1 oldal">
<mxGraphModel dx="2038" dy="647" grid="1" gridSize="10" guides="1" tooltips="1" connect="1" arrows="1" fold="1" page="1" pageScale="1" pageWidth="1200" pageHeight="1600" background="none" math="0" shadow="0">
<mxGraphModel dx="1432" dy="899" grid="1" gridSize="10" guides="1" tooltips="1" connect="1" arrows="1" fold="1" page="1" pageScale="1" pageWidth="1200" pageHeight="1600" background="none" math="0" shadow="0">
<root>
<mxCell id="0"/>
<mxCell id="1" parent="0"/>
@@ -1411,7 +1411,7 @@
<mxCell id="184" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#C23555;fontColor=#FFFFFF;strokeColor=none;" parent="183" vertex="1">
<mxGeometry width="280" height="250" as="geometry"/>
</mxCell>
<mxCell id="185" value="&lt;font face=&quot;Klavika&quot; style=&quot;font-size: 17px;&quot;&gt;Cluster&lt;/font&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" parent="183" vertex="1">
<mxCell id="185" value="&lt;font face=&quot;Klavika&quot; style=&quot;font-size: 17px;&quot;&gt;Block&lt;/font&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" parent="183" vertex="1">
<mxGeometry x="110" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="186" value="" style="group;fontColor=#FFFFFF;" parent="183" vertex="1" connectable="0">
@@ -1591,7 +1591,7 @@
<mxCell id="2939" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#585556;fontColor=#FFFFFF;strokeColor=none;" parent="2938" vertex="1">
<mxGeometry width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="2940" value="Block" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="2938" vertex="1">
<mxCell id="2940" value="Warp" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="2938" vertex="1">
<mxGeometry x="15" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="2941" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" parent="2938" edge="1">
@@ -1762,7 +1762,7 @@
<mxCell id="2972" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#585556;fontColor=#FFFFFF;strokeColor=none;" parent="2971" vertex="1">
<mxGeometry width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="2973" value="Block" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="2971" vertex="1">
<mxCell id="2973" value="Warp" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="2971" vertex="1">
<mxGeometry x="15" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="2974" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" parent="2971" edge="1">
@@ -1876,7 +1876,7 @@
<mxCell id="209" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#585556;fontColor=#FFFFFF;strokeColor=none;" parent="208" vertex="1">
<mxGeometry width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="210" value="Block" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="208" vertex="1">
<mxCell id="210" value="Warp" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="208" vertex="1">
<mxGeometry x="15" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="211" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" parent="208" edge="1">
@@ -2047,7 +2047,7 @@
<mxCell id="2906" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#585556;fontColor=#FFFFFF;strokeColor=none;" parent="2905" vertex="1">
<mxGeometry width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="2907" value="Block" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="2905" vertex="1">
<mxCell id="2907" value="Warp" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="2905" vertex="1">
<mxGeometry x="15" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="2908" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" parent="2905" edge="1">
@@ -3490,7 +3490,7 @@
<mxCell id="4199" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#C23555;fontColor=#FFFFFF;strokeColor=none;" parent="4198" vertex="1">
<mxGeometry width="280" height="250" as="geometry"/>
</mxCell>
<mxCell id="4200" value="&lt;font face=&quot;Klavika&quot; style=&quot;font-size: 17px;&quot;&gt;Cluster&lt;/font&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" parent="4198" vertex="1">
<mxCell id="4200" value="&lt;font face=&quot;Klavika&quot; style=&quot;font-size: 17px;&quot;&gt;Block&lt;/font&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" parent="4198" vertex="1">
<mxGeometry x="110" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="4201" value="" style="group;fontColor=#FFFFFF;" parent="4198" vertex="1" connectable="0">
@@ -3670,7 +3670,7 @@
<mxCell id="4235" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#585556;fontColor=#FFFFFF;strokeColor=none;" parent="4234" vertex="1">
<mxGeometry width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="4236" value="Block" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="4234" vertex="1">
<mxCell id="4236" value="Warp" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="4234" vertex="1">
<mxGeometry x="15" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="4237" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" parent="4234" edge="1">
@@ -3841,7 +3841,7 @@
<mxCell id="4268" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#585556;fontColor=#FFFFFF;strokeColor=none;" parent="4267" vertex="1">
<mxGeometry width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="4269" value="Block" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="4267" vertex="1">
<mxCell id="4269" value="Warp" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="4267" vertex="1">
<mxGeometry x="15" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="4270" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" parent="4267" edge="1">
@@ -3955,7 +3955,7 @@
<mxCell id="4290" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#585556;fontColor=#FFFFFF;strokeColor=none;" parent="4289" vertex="1">
<mxGeometry width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="4291" value="Block" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="4289" vertex="1">
<mxCell id="4291" value="Warp" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="4289" vertex="1">
<mxGeometry x="15" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="4292" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" parent="4289" edge="1">
@@ -4126,7 +4126,7 @@
<mxCell id="4323" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#585556;fontColor=#FFFFFF;strokeColor=none;" parent="4322" vertex="1">
<mxGeometry width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="4324" value="Block" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="4322" vertex="1">
<mxCell id="4324" value="Warp" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="4322" vertex="1">
<mxGeometry x="15" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="4325" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" parent="4322" edge="1">
@@ -4534,7 +4534,7 @@
<mxCell id="4604" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#C23555;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="20" y="405" width="620" height="135" as="geometry"/>
</mxCell>
<mxCell id="4605" value="&lt;font face=&quot;Klavika&quot; style=&quot;font-size: 17px;&quot;&gt;Cluster&lt;/font&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxCell id="4605" value="&lt;font face=&quot;Klavika&quot; style=&quot;font-size: 17px;&quot;&gt;Block&lt;/font&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="310" y="405" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="4683" value="" style="group;fontColor=#FFFFFF;" parent="1" vertex="1" connectable="0">
@@ -4600,7 +4600,7 @@
<mxCell id="4695" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#585556;fontColor=#FFFFFF;strokeColor=none;" parent="4694" vertex="1">
<mxGeometry width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="4696" value="Block" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="4694" vertex="1">
<mxCell id="4696" value="Warp" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="4694" vertex="1">
<mxGeometry x="15" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="4697" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" parent="4694" edge="1">
@@ -4771,7 +4771,7 @@
<mxCell id="4728" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#585556;fontColor=#FFFFFF;strokeColor=none;" parent="4727" vertex="1">
<mxGeometry width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="4729" value="Block" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="4727" vertex="1">
<mxCell id="4729" value="Warp" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="4727" vertex="1">
<mxGeometry x="15" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="4730" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" parent="4727" edge="1">
@@ -4933,7 +4933,7 @@
<mxCell id="4640" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#585556;fontColor=#FFFFFF;strokeColor=none;" parent="4639" vertex="1">
<mxGeometry width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="4641" value="Block" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="4639" vertex="1">
<mxCell id="4641" value="Warp" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="4639" vertex="1">
<mxGeometry x="15" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="4642" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" parent="4639" edge="1">
@@ -4984,163 +4984,163 @@
<mxPoint x="79.86000000000013" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5154" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="1">
<mxCell id="5154" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" parent="1" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="549.86" y="481.5" as="sourcePoint"/>
<mxPoint x="549.86" y="528.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5155" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxCell id="5155" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="540" y="455" width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="5156" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="1">
<mxCell id="5156" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" parent="1" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="559.8600000000001" y="481.5" as="sourcePoint"/>
<mxPoint x="559.8600000000001" y="528.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5157" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="1">
<mxCell id="5157" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" parent="1" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="569.86" y="481.50000000000006" as="sourcePoint"/>
<mxPoint x="569.86" y="528.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5158" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="1">
<mxCell id="5158" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" parent="1" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="579.8600000000001" y="481.50000000000006" as="sourcePoint"/>
<mxPoint x="579.8600000000001" y="528.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5159" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="1">
<mxCell id="5159" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" parent="1" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="589.86" y="481.50000000000006" as="sourcePoint"/>
<mxPoint x="589.86" y="528.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5160" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="1">
<mxCell id="5160" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" parent="1" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="599.8600000000001" y="481.50000000000006" as="sourcePoint"/>
<mxPoint x="599.8600000000001" y="528.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5161" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="1">
<mxCell id="5161" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" parent="1" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="609.86" y="481.50000000000006" as="sourcePoint"/>
<mxPoint x="609.86" y="528.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5162" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="1">
<mxCell id="5162" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" parent="1" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="619.8600000000001" y="481.50000000000006" as="sourcePoint"/>
<mxPoint x="619.8600000000001" y="528.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5163" value="" style="group;fontColor=#FFFFFF;" vertex="1" connectable="0" parent="1">
<mxCell id="5163" value="" style="group;fontColor=#FFFFFF;" parent="1" vertex="1" connectable="0">
<mxGeometry x="530" y="445" width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="5164" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#523E43;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="5163">
<mxCell id="5164" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#523E43;fontColor=#FFFFFF;strokeColor=none;" parent="5163" vertex="1">
<mxGeometry width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="5165" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5163">
<mxCell id="5165" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" parent="5163" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="9.86" y="26.5" as="sourcePoint"/>
<mxPoint x="9.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5166" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5163">
<mxCell id="5166" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" parent="5163" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="19.860000000000113" y="26.5" as="sourcePoint"/>
<mxPoint x="19.860000000000113" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5167" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5163">
<mxCell id="5167" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" parent="5163" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="29.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="29.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5168" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5163">
<mxCell id="5168" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" parent="5163" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="39.86000000000011" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="39.86000000000011" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5169" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5163">
<mxCell id="5169" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" parent="5163" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="49.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="49.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5170" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5163">
<mxCell id="5170" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" parent="5163" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="59.86000000000011" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="59.86000000000011" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5171" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5163">
<mxCell id="5171" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" parent="5163" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="69.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="69.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5172" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5163">
<mxCell id="5172" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" parent="5163" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="79.86000000000013" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="79.86000000000013" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5173" value="" style="group;fontColor=#FFFFFF;" vertex="1" connectable="0" parent="1">
<mxCell id="5173" value="" style="group;fontColor=#FFFFFF;" parent="1" vertex="1" connectable="0">
<mxGeometry x="520" y="435" width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="5174" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#585556;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="5173">
<mxCell id="5174" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#585556;fontColor=#FFFFFF;strokeColor=none;" parent="5173" vertex="1">
<mxGeometry width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="5175" value="Block" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="5173">
<mxCell id="5175" value="Warp" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="5173" vertex="1">
<mxGeometry x="15" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="5176" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5173">
<mxCell id="5176" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" parent="5173" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="9.86" y="26.5" as="sourcePoint"/>
<mxPoint x="9.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5177" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5173">
<mxCell id="5177" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" parent="5173" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="19.860000000000113" y="26.5" as="sourcePoint"/>
<mxPoint x="19.860000000000113" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5178" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5173">
<mxCell id="5178" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" parent="5173" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="29.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="29.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5179" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5173">
<mxCell id="5179" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" parent="5173" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="39.86000000000011" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="39.86000000000011" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5180" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5173">
<mxCell id="5180" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" parent="5173" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="49.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="49.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5181" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5173">
<mxCell id="5181" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" parent="5173" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="59.86000000000011" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="59.86000000000011" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5182" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5173">
<mxCell id="5182" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" parent="5173" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="69.86" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="69.86" y="73.5" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="5183" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" edge="1" parent="5173">
<mxCell id="5183" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;" parent="5173" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="79.86000000000013" y="26.500000000000057" as="sourcePoint"/>
<mxPoint x="79.86000000000013" y="73.5" as="targetPoint"/>
projects/hip/docs/data/how-to/cooperative_groups/thread_hierarchy_coop_top.svg
+1
Melihat File
File diff suppressed because one or more lines are too long
Side by Side

After

Width:  |  Height:  |  Ukuran: 308 KiB

projects/hip/docs/data/how-to/hipgraph/hip_graph.drawio
+76
Melihat File
@@ -0,0 +1,76 @@
<mxfile host="app.diagrams.net" agent="Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:130.0) Gecko/20100101 Firefox/130.0" version="24.7.7">
<diagram id="zBbb_w2fufU70cdOGtND" name="1 oldal">
<mxGraphModel dx="1114" dy="694" grid="1" gridSize="10" guides="1" tooltips="1" connect="1" arrows="1" fold="1" page="1" pageScale="1" pageWidth="1200" pageHeight="1600" background="none" math="0" shadow="0">
<root>
<mxCell id="0" />
<mxCell id="1" parent="0" />
<mxCell id="5536" value="" style="rounded=0;fillColor=#5e5b61;fontColor=#FFFFFF;strokeColor=none;spacing=0;fontFamily=Segoe UI;fontSize=18;" parent="1" vertex="1">
<mxGeometry x="340" y="10" width="280" height="520" as="geometry" />
</mxCell>
<mxCell id="1Txoek2s6jAQB3cqoh21-5821" value="" style="rounded=0;fillColor=#5e5b61;fontColor=#FFFFFF;strokeColor=none;spacing=0;fontFamily=Segoe UI;fontSize=18;" parent="1" vertex="1">
<mxGeometry x="10" y="10" width="280" height="520" as="geometry" />
</mxCell>
<mxCell id="5401" value="Stream 1" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=Segoe UI;fontSize=18;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry y="10" width="320" height="30" as="geometry" />
</mxCell>
<mxCell id="1Txoek2s6jAQB3cqoh21-5820" value="Kernel B" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#c13555;fontColor=#FFFFFF;strokeColor=none;fontFamily=Segoe UI;fontSize=18;" parent="1" vertex="1">
<mxGeometry x="30" y="170" width="240" height="100" as="geometry" />
</mxCell>
<mxCell id="1Txoek2s6jAQB3cqoh21-5819" value="Stream 2" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=Segoe UI;fontSize=18;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="320" y="10" width="320" height="30" as="geometry" />
</mxCell>
<mxCell id="1Txoek2s6jAQB3cqoh21-5822" value="Kernel A" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#c13555;fontColor=#FFFFFF;strokeColor=none;fontFamily=Segoe UI;fontSize=18;" parent="1" vertex="1">
<mxGeometry x="30" y="50" width="240" height="60" as="geometry" />
</mxCell>
<mxCell id="1Txoek2s6jAQB3cqoh21-5825" value="hipDeviceSynchronize" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#c13555;fontColor=#FFFFFF;strokeColor=none;fontFamily=Segoe UI;fontSize=18;" parent="1" vertex="1">
<mxGeometry x="30" y="330" width="570" height="40" as="geometry" />
</mxCell>
<mxCell id="1Txoek2s6jAQB3cqoh21-5826" value="Kernel C" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#c13555;fontColor=#FFFFFF;strokeColor=none;fontFamily=Segoe UI;fontSize=18;" parent="1" vertex="1">
<mxGeometry x="360" y="170" width="240" height="150" as="geometry" />
</mxCell>
<mxCell id="1Txoek2s6jAQB3cqoh21-5827" value="hipDeviceSynchronize" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#c13555;fontColor=#FFFFFF;strokeColor=none;fontFamily=Segoe UI;fontSize=18;" parent="1" vertex="1">
<mxGeometry x="30" y="120" width="570" height="40" as="geometry" />
</mxCell>
<mxCell id="1Txoek2s6jAQB3cqoh21-5828" value="Kernel D" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#c13555;fontColor=#FFFFFF;strokeColor=none;fontFamily=Segoe UI;fontSize=18;" parent="1" vertex="1">
<mxGeometry x="30" y="380" width="240" height="140" as="geometry" />
</mxCell>
<mxCell id="xSuKLjqbr4XYnemY5Hsx-5818" value="&lt;font style=&quot;font-size: 18px;&quot; color=&quot;#ffffff&quot;&gt;&lt;font style=&quot;font-size: 18px;&quot;&gt;Kernel A&lt;/font&gt;&lt;br style=&quot;font-size: 18px;&quot;&gt;&lt;/font&gt;" style="ellipse;whiteSpace=wrap;html=1;aspect=fixed;fillColor=#c13555;strokeColor=default;fontFamily=Segoe UI;fontSize=18;" parent="1" vertex="1">
<mxGeometry x="840" y="10" width="120" height="120" as="geometry" />
</mxCell>
<mxCell id="xSuKLjqbr4XYnemY5Hsx-5819" value="&lt;font style=&quot;font-size: 18px;&quot; color=&quot;#ffffff&quot;&gt;&lt;font style=&quot;font-size: 18px;&quot;&gt;Kernel B&lt;/font&gt;&lt;br style=&quot;font-size: 18px;&quot;&gt;&lt;/font&gt;" style="ellipse;whiteSpace=wrap;html=1;aspect=fixed;fillColor=#c13555;strokeColor=default;fontFamily=Segoe UI;fontSize=18;" parent="1" vertex="1">
<mxGeometry x="680" y="210" width="120" height="120" as="geometry" />
</mxCell>
<mxCell id="xSuKLjqbr4XYnemY5Hsx-5820" value="&lt;font style=&quot;font-size: 18px;&quot; color=&quot;#ffffff&quot;&gt;&lt;font style=&quot;font-size: 18px;&quot;&gt;Kernel C&lt;/font&gt;&lt;br style=&quot;font-size: 18px;&quot;&gt;&lt;/font&gt;" style="ellipse;whiteSpace=wrap;html=1;aspect=fixed;fillColor=#c13555;strokeColor=default;fontFamily=Segoe UI;fontSize=18;" parent="1" vertex="1">
<mxGeometry x="1000" y="210" width="120" height="120" as="geometry" />
</mxCell>
<mxCell id="xSuKLjqbr4XYnemY5Hsx-5821" value="&lt;div style=&quot;font-size: 18px;&quot;&gt;&lt;font style=&quot;font-size: 18px;&quot; color=&quot;#ffffff&quot;&gt;&lt;font style=&quot;font-size: 18px;&quot;&gt;Kernel D&lt;/font&gt;&lt;/font&gt;&lt;/div&gt;" style="ellipse;whiteSpace=wrap;html=1;aspect=fixed;fillColor=#c13555;strokeColor=default;fontFamily=Segoe UI;fontSize=18;" parent="1" vertex="1">
<mxGeometry x="840" y="410" width="120" height="120" as="geometry" />
</mxCell>
<mxCell id="xSuKLjqbr4XYnemY5Hsx-5822" value="" style="endArrow=blockThin;html=1;rounded=0;entryX=0;entryY=0;entryDx=0;entryDy=0;exitX=1;exitY=1;exitDx=0;exitDy=0;strokeWidth=5;endFill=1;fontFamily=Segoe UI;fontSize=18;" parent="1" source="xSuKLjqbr4XYnemY5Hsx-5818" target="xSuKLjqbr4XYnemY5Hsx-5820" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="990" y="230" as="sourcePoint" />
<mxPoint x="1060" y="190" as="targetPoint" />
</mxGeometry>
</mxCell>
<mxCell id="xSuKLjqbr4XYnemY5Hsx-5823" value="" style="endArrow=blockThin;html=1;rounded=0;entryX=1;entryY=0;entryDx=0;entryDy=0;exitX=0;exitY=1;exitDx=0;exitDy=0;strokeWidth=5;endFill=1;fontFamily=Segoe UI;fontSize=18;" parent="1" source="xSuKLjqbr4XYnemY5Hsx-5818" target="xSuKLjqbr4XYnemY5Hsx-5819" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="952" y="122" as="sourcePoint" />
<mxPoint x="1028" y="198" as="targetPoint" />
</mxGeometry>
</mxCell>
<mxCell id="xSuKLjqbr4XYnemY5Hsx-5824" value="" style="endArrow=blockThin;html=1;rounded=0;entryX=0;entryY=0;entryDx=0;entryDy=0;exitX=1;exitY=1;exitDx=0;exitDy=0;strokeWidth=5;endFill=1;fontFamily=Segoe UI;fontSize=18;" parent="1" source="xSuKLjqbr4XYnemY5Hsx-5819" target="xSuKLjqbr4XYnemY5Hsx-5821" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="868" y="162" as="sourcePoint" />
<mxPoint x="792" y="238" as="targetPoint" />
</mxGeometry>
</mxCell>
<mxCell id="xSuKLjqbr4XYnemY5Hsx-5825" value="" style="endArrow=blockThin;html=1;rounded=0;strokeWidth=5;endFill=1;entryX=1;entryY=0;entryDx=0;entryDy=0;exitX=0;exitY=1;exitDx=0;exitDy=0;fontFamily=Segoe UI;fontSize=18;" parent="1" source="xSuKLjqbr4XYnemY5Hsx-5820" target="xSuKLjqbr4XYnemY5Hsx-5821" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="1088.57" y="370" as="sourcePoint" />
<mxPoint x="1000" y="420" as="targetPoint" />
</mxGeometry>
</mxCell>
</root>
</mxGraphModel>
</diagram>
</mxfile>
projects/hip/docs/data/how-to/hipgraph/hip_graph.svg
+4
Melihat File
File diff suppressed because one or more lines are too long
Side by Side

After

Width:  |  Height:  |  Ukuran: 25 KiB

projects/hip/docs/data/how-to/hipgraph/hip_graph_speedup.drawio
+162
Melihat File
@@ -0,0 +1,162 @@
<mxfile host="app.diagrams.net" agent="Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:130.0) Gecko/20100101 Firefox/130.0" version="24.7.13">
<diagram id="zBbb_w2fufU70cdOGtND" name="1 oldal">
<mxGraphModel dx="1114" dy="646" grid="1" gridSize="10" guides="1" tooltips="1" connect="1" arrows="1" fold="1" page="1" pageScale="1" pageWidth="1200" pageHeight="1600" background="none" math="0" shadow="0">
<root>
<mxCell id="0" />
<mxCell id="1" parent="0" />
<mxCell id="5818" value="" style="rounded=0;fillColor=#5E5B61;fontColor=#FFFFFF;strokeColor=none;spacing=0;fontFamily=Segoe UI;fontSize=18;" parent="1" vertex="1">
<mxGeometry width="1130" height="460" as="geometry" />
</mxCell>
<mxCell id="5401" value="Streams" style="text;strokeColor=none;fillColor=none;align=left;verticalAlign=middle;rounded=0;fontFamily=Segoe UI;fontSize=18;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="10" y="10" width="100" height="30" as="geometry" />
</mxCell>
<mxCell id="1Txoek2s6jAQB3cqoh21-5822" value="&lt;font style=&quot;font-size: 18px;&quot;&gt;kernel A&lt;/font&gt;" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#c23555;fontColor=#FFFFFF;strokeColor=default;fontFamily=Segoe UI;fontSize=18;" parent="1" vertex="1">
<mxGeometry x="390" y="130" width="90" height="40" as="geometry" />
</mxCell>
<mxCell id="1Txoek2s6jAQB3cqoh21-5827" value="kernel launch A" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#A20025;fontColor=#FFFFFF;strokeColor=default;fontFamily=Segoe UI;fontSize=18;" parent="1" vertex="1">
<mxGeometry x="200" y="70" width="190" height="40" as="geometry" />
</mxCell>
<mxCell id="TYkUcg6durY7-jI84bFK-5819" value="&lt;div style=&quot;font-size: 18px;&quot;&gt;&lt;font style=&quot;font-size: 18px;&quot;&gt;kernel B&lt;/font&gt;&lt;/div&gt;" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#c23555;fontColor=#FFFFFF;strokeColor=default;fontFamily=Segoe UI;fontSize=18;" parent="1" vertex="1">
<mxGeometry x="580" y="130" width="70" height="40" as="geometry" />
</mxCell>
<mxCell id="TYkUcg6durY7-jI84bFK-5820" value="kernel C" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#c23555;fontColor=#FFFFFF;strokeColor=default;fontFamily=Segoe UI;fontSize=18;" parent="1" vertex="1">
<mxGeometry x="770" y="130" width="90" height="40" as="geometry" />
</mxCell>
<mxCell id="TYkUcg6durY7-jI84bFK-5822" value="kernel launch B" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#A20025;fontColor=#FFFFFF;strokeColor=default;fontFamily=Segoe UI;fontSize=18;" parent="1" vertex="1">
<mxGeometry x="390" y="70" width="190" height="40" as="geometry" />
</mxCell>
<mxCell id="TYkUcg6durY7-jI84bFK-5823" value="kernel launch C" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#A20025;fontColor=#FFFFFF;strokeColor=default;fontFamily=Segoe UI;fontSize=18;" parent="1" vertex="1">
<mxGeometry x="580" y="70" width="190" height="40" as="geometry" />
</mxCell>
<mxCell id="TYkUcg6durY7-jI84bFK-5824" value="host activity" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#A20025;fontColor=#FFFFFF;strokeColor=none;fontFamily=Segoe UI;fontSize=18;" parent="1" vertex="1">
<mxGeometry x="30" y="70" width="120" height="40" as="geometry" />
</mxCell>
<mxCell id="TYkUcg6durY7-jI84bFK-5826" value="device activity" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#c23555;fontColor=#FFFFFF;strokeColor=none;fontFamily=Segoe UI;fontSize=18;" parent="1" vertex="1">
<mxGeometry x="30" y="130" width="120" height="40" as="geometry" />
</mxCell>
<mxCell id="TYkUcg6durY7-jI84bFK-5827" value="" style="endArrow=blockThin;startArrow=blockThin;html=1;rounded=0;endFill=1;startFill=1;strokeWidth=3;fontFamily=Segoe UI;fontSize=18;" parent="1" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="480" y="210" as="sourcePoint" />
<mxPoint x="580" y="210" as="targetPoint" />
<Array as="points" />
</mxGeometry>
</mxCell>
<mxCell id="TYkUcg6durY7-jI84bFK-5829" value="" style="endArrow=blockThin;startArrow=blockThin;html=1;rounded=0;exitX=1;exitY=0.5;exitDx=0;exitDy=0;endFill=1;startFill=1;strokeWidth=3;fontFamily=Segoe UI;fontSize=18;" parent="1" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="650" y="210" as="sourcePoint" />
<mxPoint x="770" y="210" as="targetPoint" />
<Array as="points" />
</mxGeometry>
</mxCell>
<mxCell id="TYkUcg6durY7-jI84bFK-5832" value="" style="endArrow=blockThin;startArrow=none;html=1;rounded=0;exitX=1;exitY=0.5;exitDx=0;exitDy=0;endFill=1;startFill=0;strokeWidth=3;entryX=0;entryY=0.75;entryDx=0;entryDy=0;fontFamily=Segoe UI;fontSize=18;" parent="1" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="200" y="270" as="sourcePoint" />
<mxPoint x="1100" y="270" as="targetPoint" />
<Array as="points" />
</mxGeometry>
</mxCell>
<mxCell id="TYkUcg6durY7-jI84bFK-5834" value="time" style="rounded=0;whiteSpace=wrap;html=1;fillColor=none;fontColor=#FFFFFF;strokeColor=none;fontFamily=Segoe UI;fontSize=18;" parent="1" vertex="1">
<mxGeometry x="1050" y="270" width="70" height="40" as="geometry" />
</mxCell>
<mxCell id="TYkUcg6durY7-jI84bFK-5835" value="" style="endArrow=none;dashed=1;html=1;rounded=0;exitX=1;exitY=1;exitDx=0;exitDy=0;strokeWidth=3;fontFamily=Segoe UI;fontSize=18;" parent="1" source="1Txoek2s6jAQB3cqoh21-5822" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="420" y="360" as="sourcePoint" />
<mxPoint x="480" y="210" as="targetPoint" />
</mxGeometry>
</mxCell>
<mxCell id="TYkUcg6durY7-jI84bFK-5836" value="" style="endArrow=none;dashed=1;html=1;rounded=0;exitX=1;exitY=1;exitDx=0;exitDy=0;strokeWidth=3;fontFamily=Segoe UI;fontSize=18;" parent="1" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="580" y="170" as="sourcePoint" />
<mxPoint x="580" y="210" as="targetPoint" />
</mxGeometry>
</mxCell>
<mxCell id="TYkUcg6durY7-jI84bFK-5837" value="" style="endArrow=none;dashed=1;html=1;rounded=0;exitX=1;exitY=1;exitDx=0;exitDy=0;strokeWidth=3;fontFamily=Segoe UI;fontSize=18;" parent="1" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="650" y="170" as="sourcePoint" />
<mxPoint x="650" y="210" as="targetPoint" />
</mxGeometry>
</mxCell>
<mxCell id="TYkUcg6durY7-jI84bFK-5838" value="" style="endArrow=none;dashed=1;html=1;rounded=0;exitX=1;exitY=1;exitDx=0;exitDy=0;strokeWidth=3;fontFamily=Segoe UI;fontSize=18;" parent="1" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="770" y="170" as="sourcePoint" />
<mxPoint x="770" y="210" as="targetPoint" />
</mxGeometry>
</mxCell>
<mxCell id="TYkUcg6durY7-jI84bFK-5839" value="kernel launch D" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#A20025;fontColor=#FFFFFF;strokeColor=default;fontFamily=Segoe UI;fontSize=18;" parent="1" vertex="1">
<mxGeometry x="770" y="70" width="180" height="40" as="geometry" />
</mxCell>
<mxCell id="TYkUcg6durY7-jI84bFK-5840" value="kernel D" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#c23555;fontColor=#FFFFFF;strokeColor=default;fontFamily=Segoe UI;fontSize=18;" parent="1" vertex="1">
<mxGeometry x="950" y="130" width="90" height="40" as="geometry" />
</mxCell>
<mxCell id="TYkUcg6durY7-jI84bFK-5841" value="" style="endArrow=blockThin;startArrow=blockThin;html=1;rounded=0;exitX=1;exitY=0.5;exitDx=0;exitDy=0;endFill=1;startFill=1;strokeWidth=3;fontFamily=Segoe UI;fontSize=18;" parent="1" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="860" y="210" as="sourcePoint" />
<mxPoint x="950" y="210" as="targetPoint" />
<Array as="points" />
</mxGeometry>
</mxCell>
<mxCell id="TYkUcg6durY7-jI84bFK-5842" value="" style="endArrow=none;dashed=1;html=1;rounded=0;exitX=1;exitY=1;exitDx=0;exitDy=0;strokeWidth=3;fontFamily=Segoe UI;fontSize=18;" parent="1" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="859" y="170" as="sourcePoint" />
<mxPoint x="859" y="210" as="targetPoint" />
</mxGeometry>
</mxCell>
<mxCell id="TYkUcg6durY7-jI84bFK-5843" value="" style="endArrow=none;dashed=1;html=1;rounded=0;exitX=1;exitY=1;exitDx=0;exitDy=0;strokeWidth=3;fontFamily=Segoe UI;fontSize=18;" parent="1" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="949" y="170" as="sourcePoint" />
<mxPoint x="949" y="210" as="targetPoint" />
</mxGeometry>
</mxCell>
<mxCell id="TYkUcg6durY7-jI84bFK-5844" value="device idling due to kernel launch congestion" style="rounded=0;whiteSpace=wrap;html=1;fillColor=none;fontColor=#FFFFFF;strokeColor=none;fontFamily=Segoe UI;fontSize=18;" parent="1" vertex="1">
<mxGeometry x="480" y="220" width="390" height="40" as="geometry" />
</mxCell>
<mxCell id="TYkUcg6durY7-jI84bFK-5845" value="kernel A" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#c23555;fontColor=#FFFFFF;strokeColor=default;fontFamily=Segoe UI;fontSize=18;" parent="1" vertex="1">
<mxGeometry x="530" y="370" width="90" height="40" as="geometry" />
</mxCell>
<mxCell id="TYkUcg6durY7-jI84bFK-5847" value="&lt;div style=&quot;font-size: 18px;&quot;&gt;kernel B&lt;/div&gt;" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#c23555;fontColor=#FFFFFF;strokeColor=default;fontFamily=Segoe UI;fontSize=18;" parent="1" vertex="1">
<mxGeometry x="620" y="370" width="70" height="40" as="geometry" />
</mxCell>
<mxCell id="TYkUcg6durY7-jI84bFK-5848" value="kernel C" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#c23555;fontColor=#FFFFFF;strokeColor=default;fontFamily=Segoe UI;fontSize=18;" parent="1" vertex="1">
<mxGeometry x="690" y="370" width="90" height="40" as="geometry" />
</mxCell>
<mxCell id="TYkUcg6durY7-jI84bFK-5849" value="graph launch" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#A20025;fontColor=#FFFFFF;strokeColor=default;fontFamily=Segoe UI;fontSize=18;" parent="1" vertex="1">
<mxGeometry x="200" y="310" width="330" height="40" as="geometry" />
</mxCell>
<mxCell id="TYkUcg6durY7-jI84bFK-5851" value="host activity" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#A20025;fontColor=#FFFFFF;strokeColor=none;fontFamily=Segoe UI;fontSize=18;" parent="1" vertex="1">
<mxGeometry x="30" y="310" width="120" height="40" as="geometry" />
</mxCell>
<mxCell id="TYkUcg6durY7-jI84bFK-5852" value="device activity" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#c23555;fontColor=#FFFFFF;strokeColor=none;fontFamily=Segoe UI;fontSize=18;" parent="1" vertex="1">
<mxGeometry x="30" y="370" width="120" height="40" as="geometry" />
</mxCell>
<mxCell id="TYkUcg6durY7-jI84bFK-5860" value="kernel D" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#c23555;fontColor=#FFFFFF;strokeColor=default;fontFamily=Segoe UI;fontSize=18;" parent="1" vertex="1">
<mxGeometry x="780" y="370" width="90" height="40" as="geometry" />
</mxCell>
<mxCell id="TYkUcg6durY7-jI84bFK-5861" value="" style="endArrow=blockThin;startArrow=blockThin;html=1;rounded=0;exitX=1;exitY=0.5;exitDx=0;exitDy=0;endFill=1;startFill=1;strokeWidth=3;fontFamily=Segoe UI;fontSize=18;" parent="1" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="870" y="450" as="sourcePoint" />
<mxPoint x="1040" y="450" as="targetPoint" />
<Array as="points" />
</mxGeometry>
</mxCell>
<mxCell id="TYkUcg6durY7-jI84bFK-5862" value="" style="endArrow=none;dashed=1;html=1;rounded=0;exitX=1;exitY=1;exitDx=0;exitDy=0;strokeWidth=3;fontFamily=Segoe UI;fontSize=18;" parent="1" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="870" y="410" as="sourcePoint" />
<mxPoint x="870" y="450" as="targetPoint" />
</mxGeometry>
</mxCell>
<mxCell id="TYkUcg6durY7-jI84bFK-5863" value="" style="endArrow=none;dashed=1;html=1;rounded=0;exitX=1;exitY=1;exitDx=0;exitDy=0;strokeWidth=3;fontFamily=Segoe UI;fontSize=18;" parent="1" source="TYkUcg6durY7-jI84bFK-5840" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="960" y="410" as="sourcePoint" />
<mxPoint x="1040" y="450" as="targetPoint" />
</mxGeometry>
</mxCell>
<mxCell id="TYkUcg6durY7-jI84bFK-5864" value="Graph" style="text;strokeColor=none;fillColor=none;align=left;verticalAlign=middle;rounded=0;fontFamily=Segoe UI;fontSize=18;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="10" y="270" width="100" height="30" as="geometry" />
</mxCell>
<mxCell id="mD1MZE_vsaNfBI7xCEMp-5818" value="speedup" style="rounded=0;whiteSpace=wrap;html=1;fillColor=none;fontColor=#FFFFFF;strokeColor=none;fontFamily=Segoe UI;fontSize=18;" parent="1" vertex="1">
<mxGeometry x="880" y="400" width="140" height="40" as="geometry" />
</mxCell>
</root>
</mxGraphModel>
</diagram>
</mxfile>
projects/hip/docs/data/how-to/hipgraph/hip_graph_speedup.svg
+4
Melihat File
File diff suppressed because one or more lines are too long
Side by Side

After

Width:  |  Height:  |  Ukuran: 203 KiB

projects/hip/docs/data/tutorial/reduction/conflict_free_reduction.drawio
+448
Melihat File
@@ -0,0 +1,448 @@
<mxfile host="65bd71144e">
<diagram id="zBbb_w2fufU70cdOGtND" name="1 oldal">
<mxGraphModel dx="1054" dy="1139" grid="1" gridSize="10" guides="1" tooltips="1" connect="1" arrows="1" fold="1" page="0" pageScale="1" pageWidth="660" pageHeight="610" background="none" math="0" shadow="0">
<root>
<mxCell id="0"/>
<mxCell id="1" parent="0"/>
<mxCell id="2" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#5E5B61;fontColor=#FFFFFF;strokeColor=none;spacing=0;" parent="1" vertex="1">
<mxGeometry width="480" height="690" as="geometry"/>
</mxCell>
<mxCell id="3" value="Bank conflict free reduction" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="210" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="5" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#C23555;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="10" y="30" width="380" height="90" as="geometry"/>
</mxCell>
<mxCell id="6" value="&lt;font face=&quot;Klavika&quot; style=&quot;font-size: 17px;&quot;&gt;Shared&lt;/font&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="190" y="30" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="142" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="20" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="147" value="5" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="27.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="305" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="142" target="339" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="450" y="50" as="sourcePoint"/>
<mxPoint x="450" y="360" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="315" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="65" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="316" value="13" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="72.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="339" value="" style="ellipse;whiteSpace=wrap;html=1;fillColor=#4F1623;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="20" y="180" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="340" value="0" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="27.5" y="190" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="342" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="110" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="343" value="8" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="117.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="344" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="315" target="347" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="540" y="50" as="sourcePoint"/>
<mxPoint x="540" y="360" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="345" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="155" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="346" value="2" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="162.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="347" value="" style="ellipse;whiteSpace=wrap;html=1;fillColor=#4F1623;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="65" y="180" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="348" value="1" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="72.5" y="190" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="350" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="200" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="351" value="7" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="207.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="352" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="342" target="355" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="630" y="50" as="sourcePoint"/>
<mxPoint x="630" y="360" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="353" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="245" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="354" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="252.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="355" value="" style="ellipse;whiteSpace=wrap;html=1;fillColor=#4F1623;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="110" y="180" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="356" value="2" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="117.5" y="190" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="358" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="290" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="359" value="23" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="297.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="360" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;" parent="1" target="363" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="175" y="110" as="sourcePoint"/>
<mxPoint x="720" y="360" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="361" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="335" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="362" value="10" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="342.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="363" value="" style="ellipse;whiteSpace=wrap;html=1;fillColor=#4F1623;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="155" y="180" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="364" value="3" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="162.5" y="190" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="366" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#C23555;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="10" y="240" width="380" height="90" as="geometry"/>
</mxCell>
<mxCell id="367" value="&lt;font face=&quot;Klavika&quot; style=&quot;font-size: 17px;&quot;&gt;Shared&lt;/font&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="190" y="240" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="368" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="20" y="270" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="369" value="7" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="27.5" y="280" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="371" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="65" y="270" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="372" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="72.5" y="280" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="376" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="110" y="270" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="377" value="23" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="117.5" y="280" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="379" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="155" y="270" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="380" value="10" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="162.5" y="280" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="384" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="200" y="270" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="385" value="7" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="207.5" y="280" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="387" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="245" y="270" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="388" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="252.5" y="280" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="392" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="290" y="270" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="393" value="23" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="297.5" y="280" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="395" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="335" y="270" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="396" value="10" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="342.5" y="280" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="400" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="339" target="368" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="50" y="110" as="sourcePoint"/>
<mxPoint x="50" y="130" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="401" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="347" target="371" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="50" y="180" as="sourcePoint"/>
<mxPoint x="50" y="220" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="402" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="355" target="376" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="60" y="190" as="sourcePoint"/>
<mxPoint x="60" y="230" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="403" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="363" target="379" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="70" y="200" as="sourcePoint"/>
<mxPoint x="70" y="240" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="451" value="" style="edgeStyle=segmentEdgeStyle;endArrow=classic;html=1;curved=0;rounded=1;endSize=8;startSize=8;entryX=0.5;entryY=0;entryDx=0;entryDy=0;strokeColor=#FFFFFF;fontColor=#FFFFFF;dashed=1;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="350" target="339" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="140" y="310" as="sourcePoint"/>
<mxPoint x="70" y="355" as="targetPoint"/>
<Array as="points">
<mxPoint x="220" y="130"/>
<mxPoint x="40" y="130"/>
</Array>
</mxGeometry>
</mxCell>
<mxCell id="452" value="" style="edgeStyle=segmentEdgeStyle;endArrow=classic;html=1;curved=0;rounded=1;endSize=8;startSize=8;entryX=0.5;entryY=0;entryDx=0;entryDy=0;strokeColor=#FFFFFF;fontColor=#FFFFFF;dashed=1;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="353" target="347" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="230" y="120" as="sourcePoint"/>
<mxPoint x="50" y="170" as="targetPoint"/>
<Array as="points">
<mxPoint x="265" y="140"/>
<mxPoint x="85" y="140"/>
</Array>
</mxGeometry>
</mxCell>
<mxCell id="453" value="" style="edgeStyle=segmentEdgeStyle;endArrow=classic;html=1;curved=0;rounded=1;endSize=8;startSize=8;strokeColor=#FFFFFF;fontColor=#FFFFFF;dashed=1;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="358" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="275" y="120" as="sourcePoint"/>
<mxPoint x="130" y="180" as="targetPoint"/>
<Array as="points">
<mxPoint x="310" y="150"/>
<mxPoint x="130" y="150"/>
</Array>
</mxGeometry>
</mxCell>
<mxCell id="454" value="" style="edgeStyle=segmentEdgeStyle;endArrow=classic;html=1;curved=0;rounded=1;endSize=8;startSize=8;strokeColor=#FFFFFF;fontColor=#FFFFFF;dashed=1;exitX=0.5;exitY=1;exitDx=0;exitDy=0;entryX=0.5;entryY=0;entryDx=0;entryDy=0;" parent="1" source="361" target="363" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="320" y="120" as="sourcePoint"/>
<mxPoint x="140" y="200" as="targetPoint"/>
<Array as="points">
<mxPoint x="355" y="160"/>
<mxPoint x="175" y="160"/>
</Array>
</mxGeometry>
</mxCell>
<mxCell id="481" value="" style="ellipse;whiteSpace=wrap;html=1;fillColor=#4F1623;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="20" y="370" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="482" value="0" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="27.5" y="380" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="483" value="" style="ellipse;whiteSpace=wrap;html=1;fillColor=#4F1623;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="65" y="370" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="484" value="1" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="72.5" y="380" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="489" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#C23555;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="10" y="430" width="380" height="90" as="geometry"/>
</mxCell>
<mxCell id="490" value="&lt;font face=&quot;Klavika&quot; style=&quot;font-size: 17px;&quot;&gt;Shared&lt;/font&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="190" y="430" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="491" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="20" y="460" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="492" value="23" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="27.5" y="470" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="493" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="65" y="460" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="494" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="72.5" y="470" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="495" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="110" y="460" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="496" value="23" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="117.5" y="470" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="497" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="155" y="460" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="498" value="10" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="162.5" y="470" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="499" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="200" y="460" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="500" value="7" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="207.5" y="470" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="501" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="245" y="460" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="502" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="252.5" y="470" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="503" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="290" y="460" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="504" value="23" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="297.5" y="470" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="505" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="335" y="460" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="506" value="10" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="342.5" y="470" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="507" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="481" target="491" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="50" y="300" as="sourcePoint"/>
<mxPoint x="50" y="320" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="508" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="483" target="493" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="50" y="370" as="sourcePoint"/>
<mxPoint x="50" y="410" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="511" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="368" target="481" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="50" y="240" as="sourcePoint"/>
<mxPoint x="50" y="280" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="512" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="371" target="483" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="60" y="250" as="sourcePoint"/>
<mxPoint x="60" y="290" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="513" value="" style="edgeStyle=segmentEdgeStyle;endArrow=classic;html=1;curved=0;rounded=1;endSize=8;startSize=8;entryX=0.5;entryY=0;entryDx=0;entryDy=0;strokeColor=#FFFFFF;fontColor=#FFFFFF;dashed=1;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="376" target="481" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="230" y="120" as="sourcePoint"/>
<mxPoint x="50" y="190" as="targetPoint"/>
<Array as="points">
<mxPoint x="130" y="340"/>
<mxPoint x="40" y="340"/>
</Array>
</mxGeometry>
</mxCell>
<mxCell id="514" value="" style="edgeStyle=segmentEdgeStyle;endArrow=classic;html=1;curved=0;rounded=1;endSize=8;startSize=8;strokeColor=#FFFFFF;fontColor=#FFFFFF;dashed=1;exitX=0.5;exitY=1;exitDx=0;exitDy=0;entryX=0.5;entryY=0;entryDx=0;entryDy=0;" parent="1" source="379" target="483" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="140" y="330" as="sourcePoint"/>
<mxPoint x="80" y="370" as="targetPoint"/>
<Array as="points">
<mxPoint x="175" y="350"/>
<mxPoint x="85" y="350"/>
</Array>
</mxGeometry>
</mxCell>
<mxCell id="516" value="" style="ellipse;whiteSpace=wrap;html=1;fillColor=#4F1623;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="20" y="530" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="517" value="0" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="27.5" y="540" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="520" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#C23555;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="10" y="590" width="380" height="90" as="geometry"/>
</mxCell>
<mxCell id="521" value="&lt;font face=&quot;Klavika&quot; style=&quot;font-size: 17px;&quot;&gt;Shared&lt;/font&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="190" y="590" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="522" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="20" y="620" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="523" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="27.5" y="630" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="524" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="65" y="620" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="525" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="72.5" y="630" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="526" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="110" y="620" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="527" value="23" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="117.5" y="630" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="528" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="155" y="620" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="529" value="10" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="162.5" y="630" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="530" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="200" y="620" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="531" value="7" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="207.5" y="630" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="532" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="245" y="620" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="533" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="252.5" y="630" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="534" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="290" y="620" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="535" value="23" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="297.5" y="630" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="536" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="335" y="620" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="537" value="10" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="342.5" y="630" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="538" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="516" target="522" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="50" y="460" as="sourcePoint"/>
<mxPoint x="50" y="480" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="540" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="491" target="516" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="50" y="430" as="sourcePoint"/>
<mxPoint x="50" y="470" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="541" value="" style="edgeStyle=segmentEdgeStyle;endArrow=classic;html=1;curved=0;rounded=1;endSize=8;startSize=8;strokeColor=#FFFFFF;fontColor=#FFFFFF;dashed=1;exitX=0.5;exitY=1;exitDx=0;exitDy=0;entryX=1;entryY=0.5;entryDx=0;entryDy=0;" parent="1" source="493" target="516" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="185" y="330" as="sourcePoint"/>
<mxPoint x="95" y="380" as="targetPoint"/>
<Array as="points">
<mxPoint x="85" y="555"/>
</Array>
</mxGeometry>
</mxCell>
<mxCell id="542" value="Thread IDs" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;whiteSpace=wrap;rotation=0;" vertex="1" parent="1">
<mxGeometry x="400" y="170" width="62.5" height="30" as="geometry"/>
</mxCell>
<mxCell id="543" value="Data snapshot" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;whiteSpace=wrap;rotation=0;" vertex="1" parent="1">
<mxGeometry x="400" y="50" width="62.5" height="50" as="geometry"/>
</mxCell>
<mxCell id="544" value="Thread IDs" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;whiteSpace=wrap;rotation=0;" vertex="1" parent="1">
<mxGeometry x="400" y="370" width="62.5" height="30" as="geometry"/>
</mxCell>
<mxCell id="545" value="Data snapshot" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;whiteSpace=wrap;rotation=0;" vertex="1" parent="1">
<mxGeometry x="400" y="260" width="62.5" height="50" as="geometry"/>
</mxCell>
<mxCell id="546" value="Thread IDs" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;whiteSpace=wrap;rotation=0;" vertex="1" parent="1">
<mxGeometry x="400" y="540" width="62.5" height="30" as="geometry"/>
</mxCell>
<mxCell id="547" value="Data snapshot" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;whiteSpace=wrap;rotation=0;" vertex="1" parent="1">
<mxGeometry x="400" y="450" width="62.5" height="50" as="geometry"/>
</mxCell>
<mxCell id="548" value="Data snapshot" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;whiteSpace=wrap;rotation=0;" vertex="1" parent="1">
<mxGeometry x="400" y="610" width="62.5" height="50" as="geometry"/>
</mxCell>
</root>
</mxGraphModel>
</diagram>
</mxfile>
projects/hip/docs/data/tutorial/reduction/conflict_free_reduction.svg
+1
Melihat File
File diff suppressed because one or more lines are too long
Side by Side

After

Width:  |  Height:  |  Ukuran: 61 KiB

projects/hip/docs/data/tutorial/reduction/foldl.drawio
+142
Melihat File
@@ -0,0 +1,142 @@
<mxfile host="65bd71144e">
<diagram id="zBbb_w2fufU70cdOGtND" name="1 oldal">
<mxGraphModel dx="992" dy="899" grid="1" gridSize="10" guides="1" tooltips="1" connect="1" arrows="1" fold="1" page="0" pageScale="1" pageWidth="660" pageHeight="610" background="none" math="0" shadow="0">
<root>
<mxCell id="0"/>
<mxCell id="1" parent="0"/>
<mxCell id="2" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#5E5B61;fontColor=#FFFFFF;strokeColor=none;spacing=0;" parent="1" vertex="1">
<mxGeometry width="480" height="680" as="geometry"/>
</mxCell>
<mxCell id="3" value="Fold-left" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="210" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="5" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#C23555;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="130" y="30" width="330" height="120" as="geometry"/>
</mxCell>
<mxCell id="6" value="&lt;font face=&quot;Klavika&quot; style=&quot;font-size: 17px;&quot;&gt;Input&lt;/font&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="265" y="30" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="107" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="360" y="60" width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="141" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="250" y="60" width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="143" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#523E43;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="20" y="160" width="220" height="120" as="geometry"/>
</mxCell>
<mxCell id="142" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="140" y="60" width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="145" value="8" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="375" y="85" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="146" value="13" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="265" y="85" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="147" value="5" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="155" y="85" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="284" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="30" y="190" width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="285" value="z" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="45" y="215" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="288" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="140" y="190" width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="289" value="5" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="155" y="215" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="290" value="&lt;font face=&quot;Klavika&quot; style=&quot;font-size: 17px;&quot;&gt;f(z,5)&lt;/font&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="100" y="160" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="291" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="142" target="288" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="184.5" y="140" as="sourcePoint"/>
<mxPoint x="185" y="180" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="292" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#523E43;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="130" y="290" width="220" height="120" as="geometry"/>
</mxCell>
<mxCell id="293" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="140" y="320" width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="294" value="5" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="155" y="345" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="295" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="250" y="320" width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="296" value="13" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="265" y="345" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="297" value="&lt;font face=&quot;Klavika&quot; style=&quot;font-size: 17px;&quot;&gt;f(f(z,5),13)&lt;/font&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="210" y="290" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="298" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="141" target="295" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="194.5" y="150" as="sourcePoint"/>
<mxPoint x="195" y="200" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="299" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#523E43;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="240" y="420" width="220" height="120" as="geometry"/>
</mxCell>
<mxCell id="300" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="250" y="450" width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="301" value="13" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="265" y="475" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="302" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="360" y="450" width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="303" value="8" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="375" y="475" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="304" value="&lt;font face=&quot;Klavika&quot; style=&quot;font-size: 17px;&quot;&gt;f(f(f(z,5),13),8)&lt;br&gt;&lt;/font&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="320" y="420" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="305" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="107" target="302" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="305" y="150" as="sourcePoint"/>
<mxPoint x="305" y="330" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="306" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="143" target="293" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="195" y="150" as="sourcePoint"/>
<mxPoint x="195" y="200" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="308" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="292" target="300" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="140" y="290" as="sourcePoint"/>
<mxPoint x="195" y="330" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="310" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#C23555;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="345" y="550" width="120" height="120" as="geometry"/>
</mxCell>
<mxCell id="311" value="&lt;font face=&quot;Klavika&quot; style=&quot;font-size: 17px;&quot;&gt;Result&lt;/font&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="375" y="550" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="312" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="360" y="580" width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="313" value="13" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="375" y="605" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="314" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0;entryY=0;entryDx=0;entryDy=0;" parent="1" target="312" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="350" y="540" as="sourcePoint"/>
<mxPoint x="305" y="460" as="targetPoint"/>
</mxGeometry>
</mxCell>
</root>
</mxGraphModel>
</diagram>
</mxfile>
projects/hip/docs/data/tutorial/reduction/foldl.svg
+1
Melihat File
File diff suppressed because one or more lines are too long
Side by Side

After

Width:  |  Height:  |  Ukuran: 20 KiB

projects/hip/docs/data/tutorial/reduction/naive_reduction.drawio
+442
Melihat File
@@ -0,0 +1,442 @@
<mxfile host="65bd71144e">
<diagram id="zBbb_w2fufU70cdOGtND" name="1 oldal">
<mxGraphModel dx="1054" dy="1139" grid="1" gridSize="10" guides="1" tooltips="1" connect="1" arrows="1" fold="1" page="0" pageScale="1" pageWidth="660" pageHeight="610" background="none" math="0" shadow="0">
<root>
<mxCell id="0"/>
<mxCell id="1" parent="0"/>
<mxCell id="2" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#5E5B61;fontColor=#FFFFFF;strokeColor=none;spacing=0;" parent="1" vertex="1">
<mxGeometry width="480" height="610" as="geometry"/>
</mxCell>
<mxCell id="3" value="Naive Shared Reduction" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="210" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="5" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#C23555;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="10" y="30" width="380" height="90" as="geometry"/>
</mxCell>
<mxCell id="6" value="&lt;font face=&quot;Klavika&quot; style=&quot;font-size: 17px;&quot;&gt;Shared&lt;/font&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="145" y="30" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="142" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="20" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="147" value="5" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="27.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="305" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="142" target="339" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="450" y="50" as="sourcePoint"/>
<mxPoint x="450" y="360" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="315" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="65" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="316" value="13" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="72.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="339" value="" style="ellipse;whiteSpace=wrap;html=1;fillColor=#4F1623;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="20" y="130" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="340" value="0" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="27.5" y="140" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="341" value="" style="edgeStyle=segmentEdgeStyle;endArrow=classic;html=1;curved=0;rounded=1;endSize=8;startSize=8;entryX=1;entryY=0.5;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;strokeColor=#FFFFFF;fontColor=#FFFFFF;dashed=1;" parent="1" source="315" target="339" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="270" y="260" as="sourcePoint"/>
<mxPoint x="320" y="210" as="targetPoint"/>
<Array as="points">
<mxPoint x="85" y="155"/>
</Array>
</mxGeometry>
</mxCell>
<mxCell id="342" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="110" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="343" value="8" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="117.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="344" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="342" target="347" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="540" y="50" as="sourcePoint"/>
<mxPoint x="540" y="360" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="345" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="155" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="346" value="2" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="162.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="347" value="" style="ellipse;whiteSpace=wrap;html=1;fillColor=#4F1623;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="110" y="130" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="348" value="2" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="117.5" y="140" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="349" value="" style="edgeStyle=segmentEdgeStyle;endArrow=classic;html=1;curved=0;rounded=1;endSize=8;startSize=8;entryX=1;entryY=0.5;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;strokeColor=#FFFFFF;fontColor=#FFFFFF;dashed=1;" parent="1" source="345" target="347" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="360" y="260" as="sourcePoint"/>
<mxPoint x="410" y="210" as="targetPoint"/>
<Array as="points">
<mxPoint x="175" y="155"/>
</Array>
</mxGeometry>
</mxCell>
<mxCell id="350" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="200" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="351" value="7" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="207.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="352" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="350" target="355" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="630" y="50" as="sourcePoint"/>
<mxPoint x="630" y="360" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="353" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="245" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="354" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="252.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="355" value="" style="ellipse;whiteSpace=wrap;html=1;fillColor=#4F1623;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="200" y="130" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="356" value="4" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="207.5" y="140" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="357" value="" style="edgeStyle=segmentEdgeStyle;endArrow=classic;html=1;curved=0;rounded=1;endSize=8;startSize=8;entryX=1;entryY=0.5;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;strokeColor=#FFFFFF;fontColor=#FFFFFF;dashed=1;" parent="1" source="353" target="355" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="450" y="260" as="sourcePoint"/>
<mxPoint x="500" y="210" as="targetPoint"/>
<Array as="points">
<mxPoint x="265" y="155"/>
</Array>
</mxGeometry>
</mxCell>
<mxCell id="358" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="290" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="359" value="23" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="297.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="360" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="358" target="363" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="720" y="50" as="sourcePoint"/>
<mxPoint x="720" y="360" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="361" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="335" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="362" value="10" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="342.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="363" value="" style="ellipse;whiteSpace=wrap;html=1;fillColor=#4F1623;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="290" y="130" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="364" value="6" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="297.5" y="140" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="365" value="" style="edgeStyle=segmentEdgeStyle;endArrow=classic;html=1;curved=0;rounded=1;endSize=8;startSize=8;entryX=1;entryY=0.5;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;strokeColor=#FFFFFF;fontColor=#FFFFFF;dashed=1;" parent="1" source="361" target="363" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="540" y="260" as="sourcePoint"/>
<mxPoint x="590" y="210" as="targetPoint"/>
<Array as="points">
<mxPoint x="355" y="155"/>
</Array>
</mxGeometry>
</mxCell>
<mxCell id="366" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#C23555;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="10" y="190" width="380" height="90" as="geometry"/>
</mxCell>
<mxCell id="367" value="&lt;font face=&quot;Klavika&quot; style=&quot;font-size: 17px;&quot;&gt;Shared&lt;/font&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="145" y="190" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="368" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="20" y="220" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="369" value="13" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="27.5" y="230" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="370" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="368" target="373" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="450" y="210" as="sourcePoint"/>
<mxPoint x="450" y="520" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="371" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="65" y="220" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="372" value="13" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="72.5" y="230" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="373" value="" style="ellipse;whiteSpace=wrap;html=1;fillColor=#4F1623;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="20" y="290" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="374" value="0" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="27.5" y="300" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="375" value="" style="edgeStyle=segmentEdgeStyle;endArrow=classic;html=1;curved=0;rounded=1;endSize=8;startSize=8;entryX=1;entryY=0.5;entryDx=0;entryDy=0;strokeColor=#FFFFFF;fontColor=#FFFFFF;dashed=1;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="376" target="373" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="270" y="420" as="sourcePoint"/>
<mxPoint x="320" y="370" as="targetPoint"/>
<Array as="points">
<mxPoint x="130" y="315"/>
</Array>
</mxGeometry>
</mxCell>
<mxCell id="376" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="110" y="220" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="377" value="8" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="117.5" y="230" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="379" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="155" y="220" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="380" value="2" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="162.5" y="230" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="384" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="200" y="220" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="385" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="207.5" y="230" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="386" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="384" target="389" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="630" y="210" as="sourcePoint"/>
<mxPoint x="630" y="520" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="387" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="245" y="220" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="388" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="252.5" y="230" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="389" value="" style="ellipse;whiteSpace=wrap;html=1;fillColor=#4F1623;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="200" y="290" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="390" value="4" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="207.5" y="300" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="391" value="" style="edgeStyle=segmentEdgeStyle;endArrow=classic;html=1;curved=0;rounded=1;endSize=8;startSize=8;entryX=1;entryY=0.5;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;strokeColor=#FFFFFF;fontColor=#FFFFFF;dashed=1;" parent="1" source="392" target="389" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="450" y="420" as="sourcePoint"/>
<mxPoint x="500" y="370" as="targetPoint"/>
<Array as="points">
<mxPoint x="310" y="315"/>
</Array>
</mxGeometry>
</mxCell>
<mxCell id="392" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="290" y="220" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="393" value="23" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="297.5" y="230" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="395" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="335" y="220" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="396" value="10" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="342.5" y="230" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="400" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="339" target="368" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="50" y="120" as="sourcePoint"/>
<mxPoint x="50" y="140" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="401" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="347" target="376" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="50" y="190" as="sourcePoint"/>
<mxPoint x="50" y="230" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="402" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="355" target="384" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="60" y="200" as="sourcePoint"/>
<mxPoint x="60" y="240" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="403" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="363" target="392" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="70" y="210" as="sourcePoint"/>
<mxPoint x="70" y="250" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="404" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#C23555;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="10" y="350" width="380" height="90" as="geometry"/>
</mxCell>
<mxCell id="405" value="&lt;font face=&quot;Klavika&quot; style=&quot;font-size: 17px;&quot;&gt;Shared&lt;/font&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="145" y="350" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="406" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="20" y="380" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="407" value="13" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="27.5" y="390" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="408" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="406" target="411" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="450" y="370" as="sourcePoint"/>
<mxPoint x="450" y="680" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="409" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="65" y="380" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="410" value="13" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="72.5" y="390" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="411" value="" style="ellipse;whiteSpace=wrap;html=1;fillColor=#4F1623;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="20" y="450" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="412" value="0" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="27.5" y="460" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="413" value="" style="edgeStyle=segmentEdgeStyle;endArrow=classic;html=1;curved=0;rounded=1;endSize=8;startSize=8;entryX=1;entryY=0.5;entryDx=0;entryDy=0;strokeColor=#FFFFFF;fontColor=#FFFFFF;dashed=1;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="418" target="411" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="270" y="580" as="sourcePoint"/>
<mxPoint x="320" y="530" as="targetPoint"/>
<Array as="points">
<mxPoint x="220" y="475"/>
</Array>
</mxGeometry>
</mxCell>
<mxCell id="414" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="110" y="380" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="415" value="8" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="117.5" y="390" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="416" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="155" y="380" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="417" value="2" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="162.5" y="390" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="418" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="200" y="380" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="419" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="207.5" y="390" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="421" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="245" y="380" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="422" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="252.5" y="390" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="426" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="290" y="380" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="427" value="23" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="297.5" y="390" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="428" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="335" y="380" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="429" value="10" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="342.5" y="390" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="430" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="373" target="406" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="50" y="190" as="sourcePoint"/>
<mxPoint x="50" y="230" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="431" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="389" target="418" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="60" y="200" as="sourcePoint"/>
<mxPoint x="60" y="240" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="432" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#C23555;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="10" y="510" width="380" height="90" as="geometry"/>
</mxCell>
<mxCell id="433" value="&lt;font face=&quot;Klavika&quot; style=&quot;font-size: 17px;&quot;&gt;Shared&lt;/font&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="145" y="510" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="434" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="20" y="540" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="435" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="27.5" y="550" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="436" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="411" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="450" y="530" as="sourcePoint"/>
<mxPoint x="40" y="540" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="437" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="65" y="540" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="438" value="13" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="72.5" y="550" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="439" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="110" y="540" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="440" value="8" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="117.5" y="550" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="441" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="155" y="540" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="442" value="2" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="162.5" y="550" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="443" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="200" y="540" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="444" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="207.5" y="550" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="445" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="245" y="540" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="446" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="252.5" y="550" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="447" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="290" y="540" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="448" value="23" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="297.5" y="550" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="449" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="335" y="540" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="450" value="10" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="342.5" y="550" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="451" value="Thread IDs" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;whiteSpace=wrap;rotation=0;" vertex="1" parent="1">
<mxGeometry x="400" y="140" width="62.5" height="30" as="geometry"/>
</mxCell>
<mxCell id="454" value="Data snapshot" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;whiteSpace=wrap;rotation=0;" vertex="1" parent="1">
<mxGeometry x="400" y="50" width="62.5" height="50" as="geometry"/>
</mxCell>
<mxCell id="460" value="Thread IDs" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;whiteSpace=wrap;rotation=0;" vertex="1" parent="1">
<mxGeometry x="400" y="300" width="62.5" height="30" as="geometry"/>
</mxCell>
<mxCell id="461" value="Data snapshot" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;whiteSpace=wrap;rotation=0;" vertex="1" parent="1">
<mxGeometry x="400" y="210" width="62.5" height="50" as="geometry"/>
</mxCell>
<mxCell id="462" value="Thread IDs" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;whiteSpace=wrap;rotation=0;" vertex="1" parent="1">
<mxGeometry x="400" y="460" width="62.5" height="30" as="geometry"/>
</mxCell>
<mxCell id="463" value="Data snapshot" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;whiteSpace=wrap;rotation=0;" vertex="1" parent="1">
<mxGeometry x="400" y="370" width="62.5" height="50" as="geometry"/>
</mxCell>
<mxCell id="464" value="Data snapshot" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;whiteSpace=wrap;rotation=0;" vertex="1" parent="1">
<mxGeometry x="400" y="530" width="62.5" height="50" as="geometry"/>
</mxCell>
</root>
</mxGraphModel>
</diagram>
</mxfile>
projects/hip/docs/data/tutorial/reduction/naive_reduction.svg
+1
Melihat File
File diff suppressed because one or more lines are too long
Side by Side

After

Width:  |  Height:  |  Ukuran: 61 KiB

projects/hip/docs/data/tutorial/reduction/parallel_foldl.drawio
+142
Melihat File
@@ -0,0 +1,142 @@
<mxfile host="65bd71144e">
<diagram id="zBbb_w2fufU70cdOGtND" name="1 oldal">
<mxGraphModel dx="992" dy="899" grid="1" gridSize="10" guides="1" tooltips="1" connect="1" arrows="1" fold="1" page="0" pageScale="1" pageWidth="660" pageHeight="610" background="none" math="0" shadow="0">
<root>
<mxCell id="0"/>
<mxCell id="1" parent="0"/>
<mxCell id="2" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#5E5B61;fontColor=#FFFFFF;strokeColor=none;spacing=0;" parent="1" vertex="1">
<mxGeometry width="490" height="550" as="geometry"/>
</mxCell>
<mxCell id="3" value="Parallel Reduce" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="205" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="5" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#C23555;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="70" y="30" width="330" height="120" as="geometry"/>
</mxCell>
<mxCell id="6" value="&lt;font face=&quot;Klavika&quot; style=&quot;font-size: 17px;&quot;&gt;Input&lt;/font&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="205" y="30" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="107" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="300" y="60" width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="141" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="190" y="60" width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="143" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#523E43;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="10" y="160" width="220" height="120" as="geometry"/>
</mxCell>
<mxCell id="142" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="80" y="60" width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="145" value="8" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="315" y="85" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="146" value="13" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="205" y="85" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="147" value="5" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="95" y="85" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="284" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="20" y="190" width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="285" value="z" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="35" y="215" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="288" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="130" y="190" width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="289" value="5" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="145" y="215" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="290" value="&lt;font face=&quot;Klavika&quot; style=&quot;font-size: 17px;&quot;&gt;f(z,5)&lt;/font&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="90" y="160" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="291" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="142" target="288" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="124.5" y="140" as="sourcePoint"/>
<mxPoint x="125" y="180" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="292" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#523E43;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="240" y="160" width="220" height="120" as="geometry"/>
</mxCell>
<mxCell id="293" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="250" y="190" width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="294" value="13" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="265" y="215" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="295" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="360" y="190" width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="296" value="8" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="375" y="215" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="297" value="&lt;font face=&quot;Klavika&quot; style=&quot;font-size: 17px;&quot;&gt;f(13,8)&lt;/font&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="320" y="160" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="299" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#523E43;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="125" y="290" width="220" height="120" as="geometry"/>
</mxCell>
<mxCell id="300" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="135" y="320" width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="301" value="5" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="150" y="345" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="302" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="245" y="320" width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="303" value="13" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="260" y="345" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="304" value="&lt;font face=&quot;Klavika&quot; style=&quot;font-size: 17px;&quot;&gt;f(f(z,5),f(13,8))&lt;br&gt;&lt;/font&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="205" y="290" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="310" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#C23555;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="175" y="420" width="120" height="120" as="geometry"/>
</mxCell>
<mxCell id="311" value="&lt;font face=&quot;Klavika&quot; style=&quot;font-size: 17px;&quot;&gt;Result&lt;/font&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="205" y="420" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="312" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="190" y="450" width="90" height="80" as="geometry"/>
</mxCell>
<mxCell id="313" value="13" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="205" y="475" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="314" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="107" target="295" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="135" y="150" as="sourcePoint"/>
<mxPoint x="185" y="200" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="315" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="141" target="293" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="355" y="150" as="sourcePoint"/>
<mxPoint x="415" y="200" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="316" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="143" target="300" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="135" y="150" as="sourcePoint"/>
<mxPoint x="185" y="200" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="317" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="292" target="302" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="145" y="160" as="sourcePoint"/>
<mxPoint x="195" y="210" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="318" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0;entryY=0;entryDx=0;entryDy=0;exitX=0.25;exitY=1;exitDx=0;exitDy=0;" parent="1" source="299" target="312" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="130" y="290" as="sourcePoint"/>
<mxPoint x="190" y="330" as="targetPoint"/>
</mxGeometry>
</mxCell>
</root>
</mxGraphModel>
</diagram>
</mxfile>
projects/hip/docs/data/tutorial/reduction/parallel_foldl.svg
+1
Melihat File
File diff suppressed because one or more lines are too long
Side by Side

After

Width:  |  Height:  |  Ukuran: 20 KiB

projects/hip/docs/data/tutorial/reduction/reduced_divergence_reduction.drawio
+442
Melihat File
@@ -0,0 +1,442 @@
<mxfile host="65bd71144e">
<diagram id="zBbb_w2fufU70cdOGtND" name="1 oldal">
<mxGraphModel dx="1054" dy="1139" grid="1" gridSize="10" guides="1" tooltips="1" connect="1" arrows="1" fold="1" page="0" pageScale="1" pageWidth="660" pageHeight="610" background="none" math="0" shadow="0">
<root>
<mxCell id="0"/>
<mxCell id="1" parent="0"/>
<mxCell id="450" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#5E5B61;fontColor=#FFFFFF;strokeColor=none;spacing=0;" vertex="1" parent="1">
<mxGeometry width="480" height="610" as="geometry"/>
</mxCell>
<mxCell id="451" value="Reduced Divergence Reduction" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="210" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="452" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#C23555;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="10" y="30" width="380" height="90" as="geometry"/>
</mxCell>
<mxCell id="453" value="&lt;font face=&quot;Klavika&quot; style=&quot;font-size: 17px;&quot;&gt;Shared&lt;/font&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="145" y="30" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="454" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="20" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="455" value="5" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="27.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="456" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" edge="1" parent="1" source="454" target="459">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="450" y="50" as="sourcePoint"/>
<mxPoint x="450" y="360" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="457" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="65" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="458" value="13" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="72.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="459" value="" style="ellipse;whiteSpace=wrap;html=1;fillColor=#4F1623;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="20" y="130" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="460" value="0" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="27.5" y="140" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="461" value="" style="edgeStyle=segmentEdgeStyle;endArrow=classic;html=1;curved=0;rounded=1;endSize=8;startSize=8;entryX=1;entryY=0.5;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;strokeColor=#FFFFFF;fontColor=#FFFFFF;dashed=1;" edge="1" parent="1" source="457" target="459">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="270" y="260" as="sourcePoint"/>
<mxPoint x="320" y="210" as="targetPoint"/>
<Array as="points">
<mxPoint x="85" y="155"/>
</Array>
</mxGeometry>
</mxCell>
<mxCell id="462" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="110" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="463" value="8" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="117.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="464" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" edge="1" parent="1" source="462" target="467">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="540" y="50" as="sourcePoint"/>
<mxPoint x="540" y="360" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="465" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="155" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="466" value="2" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="162.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="467" value="" style="ellipse;whiteSpace=wrap;html=1;fillColor=#4F1623;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="110" y="130" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="468" value="1" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="117.5" y="140" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="469" value="" style="edgeStyle=segmentEdgeStyle;endArrow=classic;html=1;curved=0;rounded=1;endSize=8;startSize=8;entryX=1;entryY=0.5;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;strokeColor=#FFFFFF;fontColor=#FFFFFF;dashed=1;" edge="1" parent="1" source="465" target="467">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="360" y="260" as="sourcePoint"/>
<mxPoint x="410" y="210" as="targetPoint"/>
<Array as="points">
<mxPoint x="175" y="155"/>
</Array>
</mxGeometry>
</mxCell>
<mxCell id="470" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="200" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="471" value="7" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="207.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="472" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" edge="1" parent="1" source="470" target="475">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="630" y="50" as="sourcePoint"/>
<mxPoint x="630" y="360" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="473" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="245" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="474" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="252.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="475" value="" style="ellipse;whiteSpace=wrap;html=1;fillColor=#4F1623;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="200" y="130" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="476" value="2" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="207.5" y="140" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="477" value="" style="edgeStyle=segmentEdgeStyle;endArrow=classic;html=1;curved=0;rounded=1;endSize=8;startSize=8;entryX=1;entryY=0.5;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;strokeColor=#FFFFFF;fontColor=#FFFFFF;dashed=1;" edge="1" parent="1" source="473" target="475">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="450" y="260" as="sourcePoint"/>
<mxPoint x="500" y="210" as="targetPoint"/>
<Array as="points">
<mxPoint x="265" y="155"/>
</Array>
</mxGeometry>
</mxCell>
<mxCell id="478" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="290" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="479" value="23" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="297.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="480" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" edge="1" parent="1" source="478" target="483">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="720" y="50" as="sourcePoint"/>
<mxPoint x="720" y="360" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="481" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="335" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="482" value="10" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="342.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="483" value="" style="ellipse;whiteSpace=wrap;html=1;fillColor=#4F1623;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="290" y="130" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="484" value="3" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="297.5" y="140" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="485" value="" style="edgeStyle=segmentEdgeStyle;endArrow=classic;html=1;curved=0;rounded=1;endSize=8;startSize=8;entryX=1;entryY=0.5;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;strokeColor=#FFFFFF;fontColor=#FFFFFF;dashed=1;" edge="1" parent="1" source="481" target="483">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="540" y="260" as="sourcePoint"/>
<mxPoint x="590" y="210" as="targetPoint"/>
<Array as="points">
<mxPoint x="355" y="155"/>
</Array>
</mxGeometry>
</mxCell>
<mxCell id="486" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#C23555;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="10" y="190" width="380" height="90" as="geometry"/>
</mxCell>
<mxCell id="487" value="&lt;font face=&quot;Klavika&quot; style=&quot;font-size: 17px;&quot;&gt;Shared&lt;/font&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="145" y="190" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="488" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="20" y="220" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="489" value="13" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="27.5" y="230" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="490" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" edge="1" parent="1" source="488" target="493">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="450" y="210" as="sourcePoint"/>
<mxPoint x="450" y="520" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="491" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="65" y="220" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="492" value="13" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="72.5" y="230" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="493" value="" style="ellipse;whiteSpace=wrap;html=1;fillColor=#4F1623;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="20" y="290" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="494" value="0" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="27.5" y="300" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="495" value="" style="edgeStyle=segmentEdgeStyle;endArrow=classic;html=1;curved=0;rounded=1;endSize=8;startSize=8;entryX=1;entryY=0.5;entryDx=0;entryDy=0;strokeColor=#FFFFFF;fontColor=#FFFFFF;dashed=1;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" edge="1" parent="1" source="496" target="493">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="270" y="420" as="sourcePoint"/>
<mxPoint x="320" y="370" as="targetPoint"/>
<Array as="points">
<mxPoint x="130" y="315"/>
</Array>
</mxGeometry>
</mxCell>
<mxCell id="496" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="110" y="220" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="497" value="8" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="117.5" y="230" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="498" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="155" y="220" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="499" value="2" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="162.5" y="230" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="500" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="200" y="220" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="501" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="207.5" y="230" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="502" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" edge="1" parent="1" source="500" target="505">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="630" y="210" as="sourcePoint"/>
<mxPoint x="630" y="520" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="503" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="245" y="220" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="504" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="252.5" y="230" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="505" value="" style="ellipse;whiteSpace=wrap;html=1;fillColor=#4F1623;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="200" y="290" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="506" value="1" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="207.5" y="300" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="507" value="" style="edgeStyle=segmentEdgeStyle;endArrow=classic;html=1;curved=0;rounded=1;endSize=8;startSize=8;entryX=1;entryY=0.5;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;strokeColor=#FFFFFF;fontColor=#FFFFFF;dashed=1;" edge="1" parent="1" source="508" target="505">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="450" y="420" as="sourcePoint"/>
<mxPoint x="500" y="370" as="targetPoint"/>
<Array as="points">
<mxPoint x="310" y="315"/>
</Array>
</mxGeometry>
</mxCell>
<mxCell id="508" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="290" y="220" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="509" value="23" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="297.5" y="230" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="510" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="335" y="220" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="511" value="10" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="342.5" y="230" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="512" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" edge="1" parent="1" source="459" target="488">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="50" y="120" as="sourcePoint"/>
<mxPoint x="50" y="140" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="513" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" edge="1" parent="1" source="467" target="496">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="50" y="190" as="sourcePoint"/>
<mxPoint x="50" y="230" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="514" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" edge="1" parent="1" source="475" target="500">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="60" y="200" as="sourcePoint"/>
<mxPoint x="60" y="240" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="515" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" edge="1" parent="1" source="483" target="508">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="70" y="210" as="sourcePoint"/>
<mxPoint x="70" y="250" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="516" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#C23555;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="10" y="350" width="380" height="90" as="geometry"/>
</mxCell>
<mxCell id="517" value="&lt;font face=&quot;Klavika&quot; style=&quot;font-size: 17px;&quot;&gt;Shared&lt;/font&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="145" y="350" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="518" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="20" y="380" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="519" value="13" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="27.5" y="390" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="520" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" edge="1" parent="1" source="518" target="523">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="450" y="370" as="sourcePoint"/>
<mxPoint x="450" y="680" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="521" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="65" y="380" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="522" value="13" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="72.5" y="390" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="523" value="" style="ellipse;whiteSpace=wrap;html=1;fillColor=#4F1623;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="20" y="450" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="524" value="0" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="27.5" y="460" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="525" value="" style="edgeStyle=segmentEdgeStyle;endArrow=classic;html=1;curved=0;rounded=1;endSize=8;startSize=8;entryX=1;entryY=0.5;entryDx=0;entryDy=0;strokeColor=#FFFFFF;fontColor=#FFFFFF;dashed=1;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" edge="1" parent="1" source="530" target="523">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="270" y="580" as="sourcePoint"/>
<mxPoint x="320" y="530" as="targetPoint"/>
<Array as="points">
<mxPoint x="220" y="475"/>
</Array>
</mxGeometry>
</mxCell>
<mxCell id="526" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="110" y="380" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="527" value="8" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="117.5" y="390" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="528" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="155" y="380" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="529" value="2" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="162.5" y="390" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="530" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="200" y="380" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="531" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="207.5" y="390" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="532" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="245" y="380" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="533" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="252.5" y="390" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="534" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="290" y="380" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="535" value="23" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="297.5" y="390" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="536" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="335" y="380" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="537" value="10" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="342.5" y="390" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="538" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" edge="1" parent="1" source="493" target="518">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="50" y="190" as="sourcePoint"/>
<mxPoint x="50" y="230" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="539" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" edge="1" parent="1" source="505" target="530">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="60" y="200" as="sourcePoint"/>
<mxPoint x="60" y="240" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="540" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#C23555;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="10" y="510" width="380" height="90" as="geometry"/>
</mxCell>
<mxCell id="541" value="&lt;font face=&quot;Klavika&quot; style=&quot;font-size: 17px;&quot;&gt;Shared&lt;/font&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="145" y="510" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="542" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="20" y="540" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="543" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="27.5" y="550" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="544" value="" style="endArrow=classic;html=1;dashed=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" edge="1" parent="1" source="523">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="450" y="530" as="sourcePoint"/>
<mxPoint x="40" y="540" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="545" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="65" y="540" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="546" value="13" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="72.5" y="550" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="547" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="110" y="540" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="548" value="8" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="117.5" y="550" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="549" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="155" y="540" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="550" value="2" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="162.5" y="550" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="551" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="200" y="540" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="552" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="207.5" y="550" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="553" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="245" y="540" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="554" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="252.5" y="550" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="555" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="290" y="540" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="556" value="23" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="297.5" y="550" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="557" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="335" y="540" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="558" value="10" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="342.5" y="550" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="559" value="Thread IDs" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;whiteSpace=wrap;rotation=0;" vertex="1" parent="1">
<mxGeometry x="400" y="140" width="62.5" height="30" as="geometry"/>
</mxCell>
<mxCell id="560" value="Data snapshot" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;whiteSpace=wrap;rotation=0;" vertex="1" parent="1">
<mxGeometry x="400" y="50" width="62.5" height="50" as="geometry"/>
</mxCell>
<mxCell id="561" value="Thread IDs" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;whiteSpace=wrap;rotation=0;" vertex="1" parent="1">
<mxGeometry x="400" y="300" width="62.5" height="30" as="geometry"/>
</mxCell>
<mxCell id="562" value="Data snapshot" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;whiteSpace=wrap;rotation=0;" vertex="1" parent="1">
<mxGeometry x="400" y="210" width="62.5" height="50" as="geometry"/>
</mxCell>
<mxCell id="563" value="Thread IDs" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;whiteSpace=wrap;rotation=0;" vertex="1" parent="1">
<mxGeometry x="400" y="460" width="62.5" height="30" as="geometry"/>
</mxCell>
<mxCell id="564" value="Data snapshot" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;whiteSpace=wrap;rotation=0;" vertex="1" parent="1">
<mxGeometry x="400" y="370" width="62.5" height="50" as="geometry"/>
</mxCell>
<mxCell id="565" value="Data snapshot" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;whiteSpace=wrap;rotation=0;" vertex="1" parent="1">
<mxGeometry x="400" y="530" width="62.5" height="50" as="geometry"/>
</mxCell>
</root>
</mxGraphModel>
</diagram>
</mxfile>
projects/hip/docs/data/tutorial/reduction/reduced_divergence_reduction.svg
+1
Melihat File
File diff suppressed because one or more lines are too long
Side by Side

After

Width:  |  Height:  |  Ukuran: 61 KiB

projects/hip/docs/data/tutorial/reduction/warp_reduction.drawio
+421
Melihat File
@@ -0,0 +1,421 @@
<mxfile host="65bd71144e">
<diagram id="zBbb_w2fufU70cdOGtND" name="1 oldal">
<mxGraphModel dx="670" dy="996" grid="1" gridSize="10" guides="1" tooltips="1" connect="1" arrows="1" fold="1" page="0" pageScale="1" pageWidth="660" pageHeight="610" background="none" math="0" shadow="0">
<root>
<mxCell id="0"/>
<mxCell id="1" parent="0"/>
<mxCell id="2" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#5E5B61;fontColor=#FFFFFF;strokeColor=none;spacing=0;" parent="1" vertex="1">
<mxGeometry width="480" height="690" as="geometry"/>
</mxCell>
<mxCell id="3" value="Warp reduction" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="210" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="5" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#523E43;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="10" y="30" width="380" height="90" as="geometry"/>
</mxCell>
<mxCell id="6" value="&lt;font face=&quot;Klavika&quot; style=&quot;font-size: 17px;&quot;&gt;Local&lt;/font&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="170" y="30" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="142" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="20" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="147" value="5" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="27.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="305" value="" style="endArrow=classic;html=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="142" target="339" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="450" y="50" as="sourcePoint"/>
<mxPoint x="450" y="360" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="315" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="65" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="316" value="13" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="72.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="339" value="" style="ellipse;whiteSpace=wrap;html=1;fillColor=#4F1623;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="20" y="180" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="340" value="0" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="27.5" y="190" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="342" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="110" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="343" value="8" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="117.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="344" value="" style="endArrow=classic;html=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="315" target="347" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="540" y="50" as="sourcePoint"/>
<mxPoint x="540" y="360" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="345" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="155" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="346" value="2" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="162.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="347" value="" style="ellipse;whiteSpace=wrap;html=1;fillColor=#4F1623;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="65" y="180" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="348" value="1" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="72.5" y="190" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="350" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="200" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="351" value="7" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="207.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="352" value="" style="endArrow=classic;html=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="342" target="355" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="630" y="50" as="sourcePoint"/>
<mxPoint x="630" y="360" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="353" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="245" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="354" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="252.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="355" value="" style="ellipse;whiteSpace=wrap;html=1;fillColor=#4F1623;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="110" y="180" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="356" value="2" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="117.5" y="190" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="358" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="290" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="359" value="23" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="297.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="360" value="" style="endArrow=classic;html=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;" parent="1" target="363" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="175" y="110" as="sourcePoint"/>
<mxPoint x="720" y="360" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="361" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="335" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="362" value="10" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="342.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="363" value="" style="ellipse;whiteSpace=wrap;html=1;fillColor=#4F1623;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="155" y="180" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="364" value="3" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="162.5" y="190" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="366" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#523E43;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="10" y="240" width="380" height="90" as="geometry"/>
</mxCell>
<mxCell id="367" value="&lt;span style=&quot;font-family: Klavika;&quot;&gt;Local&lt;/span&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="170" y="240" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="368" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="20" y="270" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="369" value="7" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="27.5" y="280" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="371" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="65" y="270" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="372" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="72.5" y="280" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="376" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="110" y="270" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="377" value="23" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="117.5" y="280" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="379" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="155" y="270" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="380" value="10" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="162.5" y="280" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="384" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="200" y="270" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="385" value="7" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="207.5" y="280" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="387" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="245" y="270" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="388" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="252.5" y="280" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="392" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="290" y="270" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="393" value="23" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="297.5" y="280" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="395" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="335" y="270" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="396" value="10" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="342.5" y="280" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="400" value="" style="endArrow=classic;html=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="339" target="368" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="50" y="110" as="sourcePoint"/>
<mxPoint x="50" y="130" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="401" value="" style="endArrow=classic;html=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="347" target="371" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="50" y="180" as="sourcePoint"/>
<mxPoint x="50" y="220" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="402" value="" style="endArrow=classic;html=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="355" target="376" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="60" y="190" as="sourcePoint"/>
<mxPoint x="60" y="230" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="403" value="" style="endArrow=classic;html=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="363" target="379" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="70" y="200" as="sourcePoint"/>
<mxPoint x="70" y="240" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="451" value="" style="endArrow=classicThin;html=1;rounded=1;endSize=8;startSize=8;entryX=0.5;entryY=0;entryDx=0;entryDy=0;strokeColor=#FFFFFF;fontColor=#FFFFFF;exitX=0.5;exitY=1;exitDx=0;exitDy=0;endFill=1;" parent="1" source="350" target="339" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="140" y="310" as="sourcePoint"/>
<mxPoint x="70" y="355" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="452" value="" style="endArrow=classicThin;html=1;rounded=1;endSize=8;startSize=8;entryX=0.5;entryY=0;entryDx=0;entryDy=0;strokeColor=#FFFFFF;fontColor=#FFFFFF;exitX=0.5;exitY=1;exitDx=0;exitDy=0;endFill=1;" parent="1" source="353" target="347" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="230" y="120" as="sourcePoint"/>
<mxPoint x="50" y="170" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="453" value="" style="endArrow=classic;html=1;rounded=1;fontColor=#FFFFFF;exitX=0.5;exitY=1;exitDx=0;exitDy=0;strokeColor=#FFFFFF;" parent="1" source="358" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="275" y="120" as="sourcePoint"/>
<mxPoint x="130" y="180" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="454" value="" style="endArrow=classicThin;html=1;rounded=1;endSize=8;startSize=8;fontColor=#FFFFFF;exitX=0.5;exitY=1;exitDx=0;exitDy=0;entryX=0.5;entryY=0;entryDx=0;entryDy=0;strokeColor=#FFFFFF;endFill=1;" parent="1" source="361" target="363" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="320" y="120" as="sourcePoint"/>
<mxPoint x="140" y="200" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="481" value="" style="ellipse;whiteSpace=wrap;html=1;fillColor=#4F1623;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="20" y="370" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="482" value="0" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="27.5" y="380" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="483" value="" style="ellipse;whiteSpace=wrap;html=1;fillColor=#4F1623;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="65" y="370" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="484" value="1" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="72.5" y="380" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="489" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#523E43;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="10" y="430" width="380" height="90" as="geometry"/>
</mxCell>
<mxCell id="490" value="&lt;span style=&quot;font-family: Klavika;&quot;&gt;Local&lt;/span&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="170" y="430" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="491" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="20" y="460" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="492" value="23" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="27.5" y="470" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="493" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="65" y="460" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="494" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="72.5" y="470" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="495" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="110" y="460" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="496" value="23" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="117.5" y="470" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="497" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="155" y="460" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="498" value="10" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="162.5" y="470" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="499" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="200" y="460" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="500" value="7" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="207.5" y="470" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="501" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="245" y="460" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="502" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="252.5" y="470" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="503" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="290" y="460" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="504" value="23" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="297.5" y="470" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="505" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="335" y="460" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="506" value="10" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="342.5" y="470" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="507" value="" style="endArrow=classic;html=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="481" target="491" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="50" y="300" as="sourcePoint"/>
<mxPoint x="50" y="320" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="508" value="" style="endArrow=classic;html=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="483" target="493" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="50" y="370" as="sourcePoint"/>
<mxPoint x="50" y="410" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="511" value="" style="endArrow=classic;html=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="368" target="481" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="50" y="240" as="sourcePoint"/>
<mxPoint x="50" y="280" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="512" value="" style="endArrow=classic;html=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="371" target="483" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="60" y="250" as="sourcePoint"/>
<mxPoint x="60" y="290" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="513" value="" style="endArrow=classicThin;html=1;rounded=1;endSize=8;startSize=8;entryX=0.5;entryY=0;entryDx=0;entryDy=0;strokeColor=#FFFFFF;fontColor=#FFFFFF;exitX=0.5;exitY=1;exitDx=0;exitDy=0;endFill=1;" parent="1" source="376" target="481" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="230" y="120" as="sourcePoint"/>
<mxPoint x="50" y="190" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="514" value="" style="endArrow=classicThin;html=1;rounded=1;endSize=8;startSize=8;strokeColor=#FFFFFF;fontColor=#FFFFFF;exitX=0.5;exitY=1;exitDx=0;exitDy=0;entryX=0.5;entryY=0;entryDx=0;entryDy=0;endFill=1;" parent="1" source="379" target="483" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="140" y="330" as="sourcePoint"/>
<mxPoint x="80" y="370" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="516" value="" style="ellipse;whiteSpace=wrap;html=1;fillColor=#4F1623;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="20" y="530" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="517" value="0" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="27.5" y="540" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="520" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#523E43;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="10" y="590" width="380" height="90" as="geometry"/>
</mxCell>
<mxCell id="521" value="&lt;span style=&quot;font-family: Klavika;&quot;&gt;Local&lt;br&gt;&lt;/span&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="170" y="590" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="522" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="20" y="620" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="523" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="27.5" y="630" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="524" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="65" y="620" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="525" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="72.5" y="630" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="526" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="110" y="620" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="527" value="23" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="117.5" y="630" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="528" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="155" y="620" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="529" value="10" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="162.5" y="630" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="530" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="200" y="620" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="531" value="7" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="207.5" y="630" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="532" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="245" y="620" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="533" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="252.5" y="630" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="534" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="290" y="620" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="535" value="23" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="297.5" y="630" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="536" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="335" y="620" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="537" value="10" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="342.5" y="630" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="538" value="" style="endArrow=classic;html=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="516" target="522" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="50" y="460" as="sourcePoint"/>
<mxPoint x="50" y="480" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="540" value="" style="endArrow=classic;html=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;" parent="1" source="491" target="516" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="50" y="430" as="sourcePoint"/>
<mxPoint x="50" y="470" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="541" value="" style="endArrow=classicThin;html=1;rounded=1;endSize=8;startSize=8;strokeColor=#FFFFFF;fontColor=#FFFFFF;exitX=0.5;exitY=1;exitDx=0;exitDy=0;entryX=0.5;entryY=0;entryDx=0;entryDy=0;endFill=1;" parent="1" source="493" target="516" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="185" y="330" as="sourcePoint"/>
<mxPoint x="95" y="380" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="542" value="Thread IDs" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;whiteSpace=wrap;rotation=0;" parent="1" vertex="1">
<mxGeometry x="400" y="170" width="62.5" height="30" as="geometry"/>
</mxCell>
<mxCell id="543" value="Data snapshot" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;whiteSpace=wrap;rotation=0;" parent="1" vertex="1">
<mxGeometry x="400" y="50" width="62.5" height="50" as="geometry"/>
</mxCell>
<mxCell id="544" value="Thread IDs" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;whiteSpace=wrap;rotation=0;" parent="1" vertex="1">
<mxGeometry x="400" y="370" width="62.5" height="30" as="geometry"/>
</mxCell>
<mxCell id="545" value="Data snapshot" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;whiteSpace=wrap;rotation=0;" parent="1" vertex="1">
<mxGeometry x="400" y="260" width="62.5" height="50" as="geometry"/>
</mxCell>
<mxCell id="546" value="Thread IDs" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;whiteSpace=wrap;rotation=0;" parent="1" vertex="1">
<mxGeometry x="400" y="540" width="62.5" height="30" as="geometry"/>
</mxCell>
<mxCell id="547" value="Data snapshot" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;whiteSpace=wrap;rotation=0;" parent="1" vertex="1">
<mxGeometry x="400" y="450" width="62.5" height="50" as="geometry"/>
</mxCell>
<mxCell id="548" value="Data snapshot" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;whiteSpace=wrap;rotation=0;" parent="1" vertex="1">
<mxGeometry x="400" y="610" width="62.5" height="50" as="geometry"/>
</mxCell>
</root>
</mxGraphModel>
</diagram>
</mxfile>
projects/hip/docs/data/tutorial/reduction/warp_reduction.svg
+2
Melihat File
File diff suppressed because one or more lines are too long
Side by Side

After

Width:  |  Height:  |  Ukuran: 60 KiB

projects/hip/docs/data/tutorial/reduction/warp_reduction_with_shared.drawio
+707
Melihat File
@@ -0,0 +1,707 @@
<mxfile host="65bd71144e">
<diagram id="zBbb_w2fufU70cdOGtND" name="1 oldal">
<mxGraphModel dx="983" dy="1266" grid="1" gridSize="10" guides="1" tooltips="1" connect="1" arrows="1" fold="1" page="0" pageScale="1" pageWidth="660" pageHeight="610" background="none" math="0" shadow="0">
<root>
<mxCell id="0"/>
<mxCell id="1" parent="0"/>
<mxCell id="903" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#363538;fontColor=#FFFFFF;strokeColor=none;spacing=0;" vertex="1" parent="1">
<mxGeometry x="-30" y="-70" width="900" height="920" as="geometry"/>
</mxCell>
<mxCell id="896" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#5E5B61;fontColor=#FFFFFF;strokeColor=none;spacing=0;" vertex="1" parent="1">
<mxGeometry x="-10" y="590" width="150" height="240" as="geometry"/>
</mxCell>
<mxCell id="2" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#5E5B61;fontColor=#FFFFFF;strokeColor=none;spacing=0;" parent="1" vertex="1">
<mxGeometry x="-10" y="-10" width="420" height="450" as="geometry"/>
</mxCell>
<mxCell id="3" value="Warp reduction" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="170" y="-10" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="5" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#523E43;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="10" y="30" width="380" height="90" as="geometry"/>
</mxCell>
<mxCell id="6" value="&lt;font face=&quot;Klavika&quot; style=&quot;font-size: 17px;&quot;&gt;Local&lt;/font&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="170" y="30" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="142" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="20" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="147" value="5" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="27.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="315" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="65" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="316" value="13" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="72.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="342" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="110" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="343" value="8" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="117.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="345" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="155" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="346" value="2" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="162.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="350" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="200" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="351" value="7" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="207.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="353" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="245" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="354" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="252.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="358" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="290" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="359" value="23" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="297.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="361" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="335" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="362" value="10" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="342.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="366" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#523E43;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="10" y="130" width="380" height="90" as="geometry"/>
</mxCell>
<mxCell id="368" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="20" y="160" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="369" value="7" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="27.5" y="170" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="371" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="65" y="160" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="372" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="72.5" y="170" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="376" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="110" y="160" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="377" value="23" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="117.5" y="170" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="379" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="155" y="160" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="380" value="10" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="162.5" y="170" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="384" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="200" y="160" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="385" value="7" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="207.5" y="170" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="387" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="245" y="160" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="388" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="252.5" y="170" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="392" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="290" y="160" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="393" value="23" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="297.5" y="170" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="395" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="335" y="160" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="396" value="10" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="342.5" y="170" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="451" value="" style="endArrow=classicThin;html=1;rounded=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;exitX=0.5;exitY=1;exitDx=0;exitDy=0;entryX=0.5;entryY=0;entryDx=0;entryDy=0;endFill=1;" parent="1" source="350" target="368" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="140" y="310" as="sourcePoint"/>
<mxPoint x="40" y="180" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="452" value="" style="endArrow=classicThin;html=1;rounded=1;entryX=0.5;entryY=0;entryDx=0;entryDy=0;strokeColor=#FFFFFF;fontColor=#FFFFFF;exitX=0.5;exitY=1;exitDx=0;exitDy=0;endFill=1;" parent="1" source="353" target="371" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="230" y="120" as="sourcePoint"/>
<mxPoint x="85" y="180" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="453" value="" style="endArrow=classicThin;html=1;rounded=1;fontColor=#FFFFFF;exitX=0.5;exitY=1;exitDx=0;exitDy=0;entryX=0.5;entryY=0;entryDx=0;entryDy=0;strokeColor=#FFFFFF;endFill=1;" parent="1" source="358" target="376" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="275" y="120" as="sourcePoint"/>
<mxPoint x="130" y="180" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="489" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#523E43;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="10" y="230" width="380" height="90" as="geometry"/>
</mxCell>
<mxCell id="490" value="&lt;span style=&quot;font-family: Klavika;&quot;&gt;Local&lt;/span&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="170" y="230" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="491" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="20" y="260" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="492" value="23" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="27.5" y="270" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="493" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="65" y="260" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="494" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="72.5" y="270" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="495" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="110" y="260" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="496" value="23" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="117.5" y="270" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="497" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="155" y="260" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="498" value="10" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="162.5" y="270" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="499" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="200" y="260" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="500" value="7" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="207.5" y="270" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="501" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="245" y="260" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="502" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="252.5" y="270" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="503" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="290" y="260" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="504" value="23" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="297.5" y="270" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="505" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="335" y="260" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="506" value="10" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="342.5" y="270" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="511" value="" style="endArrow=classicThin;html=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;endFill=1;" parent="1" source="368" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="50" y="130" as="sourcePoint"/>
<mxPoint x="40" y="260" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="512" value="" style="endArrow=classicThin;html=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;endFill=1;" parent="1" source="371" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="60" y="140" as="sourcePoint"/>
<mxPoint x="85" y="260" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="513" value="" style="endArrow=classicThin;html=1;rounded=1;entryX=0.5;entryY=0;entryDx=0;entryDy=0;strokeColor=#FFFFFF;fontColor=#FFFFFF;exitX=0.5;exitY=1;exitDx=0;exitDy=0;endFill=1;" parent="1" source="376" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="230" y="10" as="sourcePoint"/>
<mxPoint x="40" y="260" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="514" value="" style="endArrow=classicThin;html=1;rounded=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;exitX=0.5;exitY=1;exitDx=0;exitDy=0;entryX=0.5;entryY=0;entryDx=0;entryDy=0;endFill=1;" parent="1" source="379" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="140" y="220" as="sourcePoint"/>
<mxPoint x="85" y="260" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="520" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#523E43;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="10" y="330" width="380" height="90" as="geometry"/>
</mxCell>
<mxCell id="521" value="&lt;span style=&quot;font-family: Klavika;&quot;&gt;Local&lt;br&gt;&lt;/span&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="170" y="330" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="522" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="20" y="360" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="523" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="27.5" y="370" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="524" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="65" y="360" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="525" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="72.5" y="370" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="526" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="110" y="360" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="527" value="23" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="117.5" y="370" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="528" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="155" y="360" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="529" value="10" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="162.5" y="370" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="530" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="200" y="360" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="531" value="7" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="207.5" y="370" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="532" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="245" y="360" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="533" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="252.5" y="370" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="534" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="290" y="360" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="535" value="23" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="297.5" y="370" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="536" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="335" y="360" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="537" value="10" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="342.5" y="370" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="540" value="" style="endArrow=classicThin;html=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;endFill=1;" parent="1" source="491" target="522" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="50" y="230" as="sourcePoint"/>
<mxPoint x="40" y="330" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="541" value="" style="endArrow=classicThin;html=1;rounded=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;exitX=0.5;exitY=1;exitDx=0;exitDy=0;entryX=0.5;entryY=0;entryDx=0;entryDy=0;endFill=1;" parent="1" source="493" target="522" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="185" y="130" as="sourcePoint"/>
<mxPoint x="60" y="355" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="305" value="" style="endArrow=classicThin;html=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;exitX=0.5;exitY=1;exitDx=0;exitDy=0;endFill=1;" parent="1" source="142" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="450" y="50" as="sourcePoint"/>
<mxPoint x="40" y="160" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="344" value="" style="endArrow=classicThin;html=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;exitX=0.5;exitY=1;exitDx=0;exitDy=0;endFill=1;" parent="1" source="315" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="540" y="50" as="sourcePoint"/>
<mxPoint x="85" y="160" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="352" value="" style="endArrow=classicThin;html=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;exitX=0.5;exitY=1;exitDx=0;exitDy=0;endFill=1;" parent="1" source="342" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="630" y="50" as="sourcePoint"/>
<mxPoint x="130" y="160" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="360" value="" style="endArrow=classicThin;html=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;endFill=1;" parent="1" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="175" y="110" as="sourcePoint"/>
<mxPoint x="175" y="160" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="665" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#5E5B61;fontColor=#FFFFFF;strokeColor=none;spacing=0;" parent="1" vertex="1">
<mxGeometry x="430" y="-10" width="420" height="450" as="geometry"/>
</mxCell>
<mxCell id="666" value="Warp reduction" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="620" y="-10" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="667" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#523E43;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="450" y="30" width="380" height="90" as="geometry"/>
</mxCell>
<mxCell id="668" value="&lt;font face=&quot;Klavika&quot; style=&quot;font-size: 17px;&quot;&gt;Local&lt;/font&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="610" y="30" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="669" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="460" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="670" value="3" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="467.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="671" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="505" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="672" value="2" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="512.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="673" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="550" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="674" value="4" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="557.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="675" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="595" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="676" value="1" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="602.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="677" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="640" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="678" value="7" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="647.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="679" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="685" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="680" value="11" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="692.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="681" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="730" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="682" value="8" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="737.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="683" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="775" y="60" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="684" value="14" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="782.5" y="70" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="685" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#523E43;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="450" y="130" width="380" height="90" as="geometry"/>
</mxCell>
<mxCell id="687" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="460" y="160" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="688" value="10" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="467.5" y="170" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="689" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="505" y="160" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="690" value="13" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="512.5" y="170" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="691" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="550" y="160" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="692" value="12" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="557.5" y="170" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="693" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="595" y="160" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="694" value="15" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="602.5" y="170" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="695" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="640" y="160" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="696" value="7" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="647.5" y="170" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="697" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="685" y="160" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="698" value="11" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="692.5" y="170" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="699" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="730" y="160" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="700" value="8" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="737.5" y="170" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="701" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="775" y="160" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="702" value="14" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="782.5" y="170" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="703" value="" style="endArrow=classicThin;html=1;rounded=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;exitX=0.5;exitY=1;exitDx=0;exitDy=0;entryX=0.5;entryY=0;entryDx=0;entryDy=0;endFill=1;" parent="1" source="677" target="687" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="580" y="310" as="sourcePoint"/>
<mxPoint x="480" y="180" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="704" value="" style="endArrow=classicThin;html=1;rounded=1;entryX=0.5;entryY=0;entryDx=0;entryDy=0;strokeColor=#FFFFFF;fontColor=#FFFFFF;exitX=0.5;exitY=1;exitDx=0;exitDy=0;endFill=1;" parent="1" source="679" target="689" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="670" y="120" as="sourcePoint"/>
<mxPoint x="525" y="180" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="706" value="" style="endArrow=classicThin;html=1;rounded=1;fontColor=#FFFFFF;exitX=0.5;exitY=1;exitDx=0;exitDy=0;entryX=0.5;entryY=0;entryDx=0;entryDy=0;strokeColor=#FFFFFF;endFill=1;" parent="1" source="683" target="693" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="760" y="120" as="sourcePoint"/>
<mxPoint x="615" y="180" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="707" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#523E43;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="450" y="230" width="380" height="90" as="geometry"/>
</mxCell>
<mxCell id="708" value="&lt;span style=&quot;font-family: Klavika;&quot;&gt;Local&lt;/span&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="610" y="230" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="709" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="460" y="260" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="710" value="22" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="467.5" y="270" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="711" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="505" y="260" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="712" value="28" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="512.5" y="270" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="713" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="550" y="260" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="714" value="12" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="557.5" y="270" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="715" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="595" y="260" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="716" value="15" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="602.5" y="270" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="717" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="640" y="260" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="718" value="7" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="647.5" y="270" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="719" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="685" y="260" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="720" value="11" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="692.5" y="270" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="721" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="730" y="260" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="722" value="8" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="737.5" y="270" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="723" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="775" y="260" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="724" value="14" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="782.5" y="270" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="725" value="" style="endArrow=classicThin;html=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;endFill=1;" parent="1" source="687" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="490" y="130" as="sourcePoint"/>
<mxPoint x="480" y="260" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="726" value="" style="endArrow=classicThin;html=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;endFill=1;" parent="1" source="689" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="500" y="140" as="sourcePoint"/>
<mxPoint x="525" y="260" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="727" value="" style="endArrow=classicThin;html=1;rounded=1;endSize=8;startSize=8;entryX=0.5;entryY=0;entryDx=0;entryDy=0;strokeColor=#FFFFFF;fontColor=#FFFFFF;exitX=0.5;exitY=1;exitDx=0;exitDy=0;endFill=1;" parent="1" source="691" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="670" y="10" as="sourcePoint"/>
<mxPoint x="480" y="260" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="728" value="" style="endArrow=classicThin;html=1;rounded=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;exitX=0.5;exitY=1;exitDx=0;exitDy=0;entryX=0.5;entryY=0;entryDx=0;entryDy=0;endFill=1;" parent="1" source="693" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="580" y="220" as="sourcePoint"/>
<mxPoint x="525" y="260" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="729" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#523E43;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="450" y="330" width="380" height="90" as="geometry"/>
</mxCell>
<mxCell id="730" value="&lt;span style=&quot;font-family: Klavika;&quot;&gt;Local&lt;br&gt;&lt;/span&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="610" y="330" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="731" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="460" y="360" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="732" value="50" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="467.5" y="370" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="733" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="505" y="360" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="734" value="28" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="512.5" y="370" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="735" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="550" y="360" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="736" value="12" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="557.5" y="370" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="737" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="595" y="360" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="738" value="15" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="602.5" y="370" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="739" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="640" y="360" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="740" value="7" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="647.5" y="370" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="741" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="685" y="360" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="742" value="11" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="692.5" y="370" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="743" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="730" y="360" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="744" value="8" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="737.5" y="370" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="745" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" parent="1" vertex="1">
<mxGeometry x="775" y="360" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="746" value="14" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="782.5" y="370" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="747" value="" style="endArrow=classicThin;html=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;entryX=0.5;entryY=0;entryDx=0;entryDy=0;exitX=0.5;exitY=1;exitDx=0;exitDy=0;endFill=1;" parent="1" source="709" target="731" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="490" y="230" as="sourcePoint"/>
<mxPoint x="480" y="330" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="748" value="" style="endArrow=classicThin;html=1;rounded=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;exitX=0.5;exitY=1;exitDx=0;exitDy=0;entryX=0.5;entryY=0;entryDx=0;entryDy=0;endFill=1;" parent="1" source="711" target="731" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="625" y="130" as="sourcePoint"/>
<mxPoint x="500" y="355" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="749" value="" style="endArrow=classicThin;html=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;exitX=0.5;exitY=1;exitDx=0;exitDy=0;endFill=1;" parent="1" source="669" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="890" y="50" as="sourcePoint"/>
<mxPoint x="480" y="160" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="750" value="" style="endArrow=classicThin;html=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;exitX=0.5;exitY=1;exitDx=0;exitDy=0;endFill=1;" parent="1" source="671" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="980" y="50" as="sourcePoint"/>
<mxPoint x="525" y="160" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="751" value="" style="endArrow=classicThin;html=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;exitX=0.5;exitY=1;exitDx=0;exitDy=0;endFill=1;" parent="1" source="673" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="1070" y="50" as="sourcePoint"/>
<mxPoint x="570" y="160" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="752" value="" style="endArrow=classicThin;html=1;strokeColor=#FFFFFF;fontColor=#FFFFFF;endFill=1;" parent="1" edge="1">
<mxGeometry width="50" height="50" relative="1" as="geometry">
<mxPoint x="615" y="110" as="sourcePoint"/>
<mxPoint x="615" y="160" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="872" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#523E43;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="10" y="720" width="105" height="90" as="geometry"/>
</mxCell>
<mxCell id="873" value="&lt;span style=&quot;font-family: Klavika;&quot;&gt;Local&lt;/span&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="32.5" y="720" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="874" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="20" y="750" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="875" value="92" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="27.5" y="760" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="876" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="65" y="750" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="877" value="50" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="72.5" y="760" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="884" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#523E43;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="10" y="610" width="105" height="90" as="geometry"/>
</mxCell>
<mxCell id="885" value="&lt;span style=&quot;font-family: Klavika;&quot;&gt;Local&lt;/span&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="32.5" y="610" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="886" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="20" y="640" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="887" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="27.5" y="650" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="888" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="65" y="640" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="889" value="50" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="72.5" y="650" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="890" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#C23555;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="10" y="470" width="105" height="90" as="geometry"/>
</mxCell>
<mxCell id="891" value="&lt;span style=&quot;font-family: Klavika;&quot;&gt;Shared&lt;/span&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="32.5" y="470" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="900" style="edgeStyle=orthogonalEdgeStyle;html=1;exitX=0.5;exitY=1;exitDx=0;exitDy=0;dashed=1;strokeColor=#FFFFFF;endArrow=classicThin;endFill=1;" edge="1" parent="1" source="892">
<mxGeometry relative="1" as="geometry">
<mxPoint x="40" y="640" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="892" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="20" y="500" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="893" value="42" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="27.5" y="510" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="901" style="edgeStyle=orthogonalEdgeStyle;html=1;exitX=0.5;exitY=1;exitDx=0;exitDy=0;dashed=1;strokeColor=#FFFFFF;endArrow=classicThin;endFill=1;" edge="1" parent="1" source="894">
<mxGeometry relative="1" as="geometry">
<mxPoint x="85" y="640" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="894" value="" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#262626;fontColor=#FFFFFF;strokeColor=none;" vertex="1" parent="1">
<mxGeometry x="65" y="500" width="40" height="50" as="geometry"/>
</mxCell>
<mxCell id="895" value="50" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontFamily=Klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="72.5" y="510" width="25" height="30" as="geometry"/>
</mxCell>
<mxCell id="897" style="html=1;exitX=0.5;exitY=1;exitDx=0;exitDy=0;entryX=0.104;entryY=1.088;entryDx=0;entryDy=0;entryPerimeter=0;dashed=1;strokeColor=#FFFFFF;endArrow=classicThin;endFill=1;" edge="1" parent="1" source="522" target="891">
<mxGeometry relative="1" as="geometry"/>
</mxCell>
<mxCell id="898" style="edgeStyle=orthogonalEdgeStyle;html=1;exitX=0.5;exitY=1;exitDx=0;exitDy=0;entryX=0.5;entryY=0;entryDx=0;entryDy=0;dashed=1;strokeColor=#FFFFFF;endArrow=classicThin;endFill=1;" edge="1" parent="1" source="731" target="894">
<mxGeometry relative="1" as="geometry"/>
</mxCell>
<mxCell id="904" value="Warp reduction with shared memory" style="text;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;rounded=0;fontFamily=klavika;fontSize=17;fontColor=#FFFFFF;" vertex="1" parent="1">
<mxGeometry x="290" y="-60" width="270" height="30" as="geometry"/>
</mxCell>
<mxCell id="367" value="&lt;span style=&quot;font-family: Klavika;&quot;&gt;Local&lt;/span&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="170" y="130" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="686" value="&lt;span style=&quot;font-family: Klavika;&quot;&gt;Local&lt;/span&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;fontSize=17;fontColor=#FFFFFF;" parent="1" vertex="1">
<mxGeometry x="610" y="130" width="60" height="30" as="geometry"/>
</mxCell>
<mxCell id="907" style="edgeStyle=none;html=1;exitX=0.5;exitY=1;exitDx=0;exitDy=0;entryX=0.084;entryY=1.029;entryDx=0;entryDy=0;entryPerimeter=0;strokeColor=#FFFFFF;endArrow=classicThin;endFill=1;" edge="1" parent="1" source="361" target="367">
<mxGeometry relative="1" as="geometry"/>
</mxCell>
<mxCell id="908" style="edgeStyle=none;html=1;exitX=0.5;exitY=1;exitDx=0;exitDy=0;entryX=0.5;entryY=0;entryDx=0;entryDy=0;strokeColor=#FFFFFF;endArrow=classicThin;endFill=1;" edge="1" parent="1" source="681" target="691">
<mxGeometry relative="1" as="geometry"/>
</mxCell>
<mxCell id="878" style="edgeStyle=orthogonalEdgeStyle;html=1;exitX=0.5;exitY=1;exitDx=0;exitDy=0;strokeColor=#FFFFFF;endArrow=classicThin;endFill=1;" edge="1" parent="1">
<mxGeometry relative="1" as="geometry">
<mxPoint x="39.999999999999886" y="689.9999999999998" as="sourcePoint"/>
<mxPoint x="40" y="750" as="targetPoint"/>
</mxGeometry>
</mxCell>
<mxCell id="879" style="html=1;exitX=0.5;exitY=1;exitDx=0;exitDy=0;strokeColor=#FFFFFF;endArrow=classicThin;endFill=1;" edge="1" parent="1">
<mxGeometry relative="1" as="geometry">
<mxPoint x="85" y="689.9999999999998" as="sourcePoint"/>
<mxPoint x="40" y="750" as="targetPoint"/>
</mxGeometry>
</mxCell>
</root>
</mxGraphModel>
</diagram>
</mxfile>
projects/hip/docs/data/tutorial/reduction/warp_reduction_with_shared.svg
+3
Melihat File
File diff suppressed because one or more lines are too long
Side by Side

After

Width:  |  Height:  |  Ukuran: 99 KiB

projects/hip/docs/data/understand/programming_model_reference/memory_hierarchy.drawio → projects/hip/docs/data/understand/programming_model/memory_hierarchy.drawio
Melihat File
projects/hip/docs/data/understand/programming_model_reference/memory_hierarchy.svg → projects/hip/docs/data/understand/programming_model/memory_hierarchy.svg
Melihat File
Side by Side Swipe Overlay

Before

Width:  |  Height:  |  Ukuran: 34 KiB

After

Width:  |  Height:  |  Ukuran: 34 KiB

projects/hip/docs/data/understand/programming_model_reference/thread_hierarchy.drawio → projects/hip/docs/data/understand/programming_model/thread_hierarchy.drawio
Melihat File
projects/hip/docs/data/understand/programming_model_reference/thread_hierarchy.svg → projects/hip/docs/data/understand/programming_model/thread_hierarchy.svg
Melihat File
Side by Side Swipe Overlay

Before

Width:  |  Height:  |  Ukuran: 477 KiB

After

Width:  |  Height:  |  Ukuran: 477 KiB

projects/hip/docs/data/understand/programming_model_reference/thread_hierarchy_coop.svg
-1
Melihat File
File diff suppressed because one or more lines are too long
Side by Side

Before

Width:  |  Height:  |  Ukuran: 308 KiB

projects/hip/docs/data/understand/textures/border.png
TEMPAT SAMPAH
Melihat File
Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Ukuran: 1.1 KiB

projects/hip/docs/data/understand/textures/clamp.png
TEMPAT SAMPAH
Melihat File
Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Ukuran: 1.1 KiB

projects/hip/docs/data/understand/textures/linear.png
TEMPAT SAMPAH
Melihat File
Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Ukuran: 2.9 KiB

projects/hip/docs/data/understand/textures/mirror.png
TEMPAT SAMPAH
Melihat File
Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Ukuran: 1.1 KiB

projects/hip/docs/data/understand/textures/nearest.png
TEMPAT SAMPAH
Melihat File
Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Ukuran: 1.3 KiB

projects/hip/docs/data/understand/textures/original.png
TEMPAT SAMPAH
Melihat File
Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Ukuran: 401 B

projects/hip/docs/data/understand/textures/wrap.png
TEMPAT SAMPAH
Melihat File
Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Ukuran: 1.1 KiB

projects/hip/docs/data/unified_memory/um.drawio
+1878
Melihat File
File diff ditekan karena terlalu besar Load Diff
projects/hip/docs/data/unified_memory/um.svg
+4
Melihat File
File diff suppressed because one or more lines are too long
Side by Side

After

Width:  |  Height:  |  Ukuran: 338 KiB

projects/hip/docs/doxygen/Doxyfile
+18 -5
Melihat File
@@ -829,10 +829,10 @@ WARN_LOGFILE =
# spaces. See also FILE_PATTERNS and EXTENSION_MAPPING
# Note: If this tag is empty the current directory is searched.
INPUT = mainpage.md \
../../include/hip \
INPUT = ../../include/hip \
../../../clr/hipamd/include/hip/amd_detail/amd_hip_gl_interop.h \
../../../llvm-project/clang/lib/Headers/__clang_hip_math.h \
../../../clr/hipamd/include/hip/amd_detail/amd_surface_functions.h \
../../../clr/hipamd/include/hip/amd_detail/amd_hip_cooperative_groups.h \
../../../ROCR-Runtime/src/inc/hsa_ext_amd.h
# This tag can be used to specify the character encoding of the source files
@@ -2195,8 +2195,21 @@ INCLUDE_FILE_PATTERNS =
# recursively expanded use the := operator instead of the = operator.
# This tag requires that the tag ENABLE_PREPROCESSING is set to YES.
PREDEFINED = __HIP_PLATFORM_AMD__ \
__dparm(x)=
PREDEFINED = "__HIP_PLATFORM_AMD__" \
"DOXYGEN_SHOULD_INCLUDE_THIS=1" \
"DOXYGEN_SHOULD_SKIP_THIS=1" \
"__dparm(x)=" \
"__cplusplus=201103L" \
"__host__=" \
"__device__=" \
"__hip_img_chk__=" \
"__CG_QUALIFIER__=" \
"__CG_STATIC_QUALIFIER__=static" \
"_CG_STATIC_CONST_DECL_=static constexpr" \
"HIP_PUBLIC_API" \
"HIP_ENABLE_WARP_SYNC_BUILTINS" \
"__HOST_DEVICE__" \
"__forceinline__"
# If the MACRO_EXPANSION and EXPAND_ONLY_PREDEF tags are set to YES then this
# tag can be used to specify a list of macro names that should be expanded. The
projects/hip/docs/doxygen/mainpage.md
-34
Melihat File
@@ -1,34 +0,0 @@
# HIP Runtime API Reference {#mainpage}
This is the full HIP Runtime API reference. The API is organized into
[modules](modules.html) based on functionality.
## List of Modules
- @ref GlobalDefs
- @ref Driver
- @ref Device
- @ref Execution
- @ref Error
- @ref Stream
- @ref StreamM
- @ref Event
- @ref Memory
- @ref External
- @ref MemoryM
- @ref StreamO
- @ref MemoryD
- @ref PeerToPeer
- @ref Context
- @ref Module
- @ref Occupancy
- @ref Profiler
- @ref Clang
- @ref Texture
- @ref TextureD
- @ref Runtime
- @ref Callback
- @ref Graph
- @ref Virtual
- @ref GL
- @ref Surface
projects/hip/docs/environment.yml
+10
Melihat File
@@ -0,0 +1,10 @@
name: RTD
channels:
- conda-forge
- defaults
dependencies:
- python=3.10
- pip
- doxygen=1.9.8
- pip:
- -r ./sphinx/requirements.txt
projects/hip/docs/how-to/cooperative_groups.rst
+490
Melihat File
@@ -0,0 +1,490 @@
.. meta::
:description: This topic describes how to use cooperative groups in HIP
:keywords: AMD, ROCm, HIP, cooperative groups
.. _cooperative_groups_how-to:
*******************************************************************************
Cooperative groups
*******************************************************************************
Cooperative groups API is an extension to the HIP programming model, which provides developers with a flexible, dynamic grouping mechanism for the communicating threads. Cooperative groups let you define your own set of thread groups which may fit your user-cases better than those defined by the hardware. This lets you specify the level of granularity for thread communication which can lead to more efficient parallel decompositions.
The API is accessible in the ``cooperative_groups`` namespace after the ``hip_cooperative_groups.h`` is included. The header contains the following elements:
* Static functions to create groups and subgroups.
* Hardware-accelerated operations over the whole group, like shuffles.
* Data types of cooperative groups.
* Synchronize member function of the groups.
* Get group properties member functions.
Cooperative groups thread model
===============================
The thread hierarchy abstraction of cooperative groups are in :ref:`grid hierarchy <coop_thread_top_hierarchy>` and :ref:`block hierarchy <coop_thread_bottom_hierarchy>`.
.. _coop_thread_top_hierarchy:
.. figure:: ../data/how-to/cooperative_groups/thread_hierarchy_coop_top.svg
:alt: Diagram depicting nested rectangles of varying color. The outermost one
titled "Grid", inside sets of different sized rectangles layered on
one another titled "Block". Each "Block" containing sets of uniform
rectangles layered on one another titled "Warp". Each of the "Warp"
titled rectangles filled with downward pointing arrows inside.
Cooperative group thread hierarchy in grids.
The **multi grid** is an abstraction of potentially multiple simultaneous launches of the same kernel over multiple devices (Deprecated since 5.0). The **grid** in cooperative groups is a single dispatch of kernels for execution like the original grid.
.. note::
The ability to synchronize over a grid or multi grid requires the kernel to be launched using the specific cooperative groups API.
The **block** is the same as the :ref:`inherent_thread_model` block entity.
.. note::
Explicit warp-level thread handling is absent from the Cooperative groups API. In order to exploit the known hardware SIMD width on which built-in functionality translates to simpler logic, you can use the group partitioning part of the API, such as ``tiled_partition``.
.. _coop_thread_bottom_hierarchy:
.. figure:: ../data/how-to/cooperative_groups/thread_hierarchy_coop_bottom.svg
:alt: The new level between block thread and threads.
Cooperative group thread hierarchy in blocks.
The cooperative groups API introduce a new level between block thread and threads. The :ref:`thread-block tile <coop_thread_block_tile>` give the opportunity to have tiles in the thread block, while the :ref:`coalesced group <coop_coalesced_groups>` holds the active threads of the parent group. These groups further discussed in the :ref:`groups types <coop_group_types>` section.
For details on memory model, check the :ref:`memory model description <memory_hierarchy>`.
.. _coop_group_types:
Group types
===========
Group types are based on the levels of synchronization and data sharing among threads.
Thread-block group
------------------
Represents an intra-block cooperative groups type where the participating threads within the group are the same threads that participated in the currently executing ``block``.
.. code-block:: cpp
class thread_block;
Constructed via:
.. code-block:: cpp
thread_block g = this_thread_block();
The ``group_index()`` , ``thread_index()`` , ``thread_rank()`` , ``size()``, ``cg_type()``, ``is_valid()`` , ``sync()`` and ``group_dim()`` member functions are public of the thread_block class. For further details, check the :ref:`thread_block references <thread_block_ref>` .
Grid group
------------
Represents an inter-block cooperative groups type where the group's participating threads span multiple blocks running the same kernel on the same device. Use the cooperative launch API to enable synchronization across the grid group.
.. code-block:: cpp
class grid_group;
Constructed via:
.. code-block:: cpp
grid_group g = this_grid();
The ``thread_rank()`` , ``size()``, ``cg_type()``, ``is_valid()`` and ``sync()`` member functions
are public of the ``grid_group`` class. For further details, check the :ref:`grid_group references <grid_group_ref>`.
Multi-grid group
------------------
Represents an inter-device cooperative groups type where the participating threads within the group span multiple devices that run the same kernel on the devices. Use the cooperative launch API to enable synchronization across the multi-grid group.
.. code-block:: cpp
class multi_grid_group;
Constructed via:
.. code-block:: cpp
// Kernel must be launched with the cooperative multi-device API
multi_grid_group g = this_multi_grid();
The ``num_grids()`` , ``grid_rank()`` , ``thread_rank()``, ``size()``, ``cg_type()``, ``is_valid()`` ,
and ``sync()`` member functions are public of the ``multi_grid_group`` class. For
further details check the :ref:`multi_grid_group references <multi_grid_group_ref>` .
.. _coop_thread_block_tile:
Thread-block tile
------------------
This constructs a templated class derived from ``thread_group``. The template defines the tile
size of the new thread group at compile time. This group type also supports sub-wave level intrinsics.
.. code-block:: cpp
template <unsigned int Size, typename ParentT = void>
class thread_block_tile;
Constructed via:
.. code-block:: cpp
template <unsigned int Size, typename ParentT>
_CG_QUALIFIER thread_block_tile<Size, ParentT> tiled_partition(const ParentT& g)
.. note::
* Size must be a power of 2 and not larger than warp (wavefront) size.
* ``shfl()`` functions support integer or float type.
The ``thread_rank()`` , ``size()``, ``cg_type()``, ``is_valid()``, ``sync()``, ``meta_group_rank()``, ``meta_group_size()``, ``shfl()``, ``shfl_down()``, ``shfl_up()``, ``shfl_xor()``, ``ballot()``, ``any()``, ``all()``, ``match_any()`` and ``match_all()`` member functions are public of the ``thread_block_tile`` class. For further details, check the :ref:`thread_block_tile references <thread_block_tile_ref>` .
.. _coop_coalesced_groups:
Coalesced groups
------------------
Threads (64 threads on CDNA and 32 threads on RDNA) in a warp cannot execute different instructions simultaneously, so conditional branches are executed serially within the warp. When threads encounter a conditional branch, they can diverge, resulting in some threads being disabled, if they do not meet the condition to execute that branch. The active threads referred as coalesced, and coalesced group represents an active thread group within a warp.
.. note::
The NVIDIA GPU's independent thread scheduling presents the appearance that threads on different branches execute concurrently.
.. warning::
AMD GPUs do not support independent thread scheduling. Some CUDA application can rely on this feature and the ported HIP version on AMD GPUs can deadlock, when they try to make use of independent thread scheduling.
This group type also supports sub-wave level intrinsics.
.. code-block:: cpp
class coalesced_group;
Constructed via:
.. code-block:: cpp
coalesced_group active = coalesced_threads();
.. note::
``shfl()`` functions support integer or float type.
The ``thread_rank()`` , ``size()``, ``cg_type()``, ``is_valid()``, ``sync()``, ``meta_group_rank()``, ``meta_group_size()``, ``shfl()``, ``shfl_down()``, ``shfl_up()``, ``ballot()``, ``any()``, ``all()``, ``match_any()`` and ``match_all()`` member functions are public of the ``coalesced_group`` class. For more information, see :ref:`coalesced_group references <coalesced_group_ref>` .
Cooperative groups simple example
=================================
The difference to the original block model in the ``reduce_sum`` device function is the following.
.. tab-set::
.. tab-item:: Original Block
:sync: original-block
.. code-block:: cuda
__device__ int reduce_sum(int *shared, int val) {
// Thread ID
const unsigned int thread_id = threadIdx.x;
// Every iteration the number of active threads
// halves, until we processed all values
for(unsigned int i = blockDim.x / 2; i > 0; i /= 2) {
// Store value in shared memory with thread ID
shared[thread_id] = val;
// Synchronize all threads
__syncthreads();
// Active thread sum up
if(thread_id < i)
val += shared[thread_id + i];
// Synchronize all threads in the group
__syncthreads();
}
// ...
}
.. tab-item:: Cooperative groups
:sync: cooperative-groups
.. code-block:: cuda
__device__ int reduce_sum(thread_group g,
int *shared,
int val) {
// Thread ID
const unsigned int group_thread_id = g.thread_rank();
// Every iteration the number of active threads
// halves, until we processed all values
for(unsigned int i = g.size() / 2; i > 0; i /= 2) {
// Store value in shared memroy with thread ID
shared[group_thread_id] = val;
// Synchronize all threads in the group
g.sync();
// Active thread sum up
if(group_thread_id < i)
val += shared[group_thread_id + i];
// Synchronize all threads in the group
g.sync();
}
// ...
}
The ``reduce_sum()`` function call and input data initialization difference to the original block model is the following.
.. tab-set::
.. tab-item:: Original Block
:sync: original-block
.. code-block:: cuda
__global__ void sum_kernel(...) {
// ...
// Workspace array in shared memory
__shared__ unsigned int workspace[2048];
// ...
// Perform reduction
output = reduce_sum(workspace, input);
// ...
}
.. tab-item:: Cooperative groups
:sync: cooperative-groups
.. code-block:: cuda
__global__ void sum_kernel(...) {
// ...
// Workspace array in shared memory
__shared__ unsigned int workspace[2048];
// ...
// Initialize the thread_block
thread_block thread_block_group = this_thread_block();
// Perform reduction
output = reduce_sum(thread_block_group, workspace, input);
// ...
}
At the device function, the input group type is the ``thread_group``, which is the parent class of all the cooperative groups type. With this, you can write generic functions, which can work with any type of cooperative groups.
.. _coop_synchronization:
Synchronization
===============
With each group type, the synchronization requires using the correct cooperative groups launch API.
**Check the kernel launch capability**
.. tab-set::
.. tab-item:: Thread-block
:sync: thread-block
Do not need kernel launch validation.
.. tab-item:: Grid
:sync: grid
Confirm the cooperative launch capability on the single AMD GPU:
.. code-block:: cpp
int device = 0;
int supports_coop_launch = 0;
// Check support
// Use hipDeviceAttributeCooperativeMultiDeviceLaunch when launching across multiple devices
HIP_CHECK(hipGetDevice(&device));
HIP_CHECK(
hipDeviceGetAttribute(&supports_coop_launch, hipDeviceAttributeCooperativeLaunch, device));
if(!supports_coop_launch)
{
std::cout << "Skipping, device " << device << " does not support cooperative groups"
<< std::endl;
return 0;
}
.. tab-item:: Multi-grid
:sync: multi-grid
Confirm the cooperative launch capability over multiple GPUs:
.. code-block:: cpp
// Check support of cooperative groups
std::vector<int> deviceIDs;
for(int deviceID = 0; deviceID < device_count; deviceID++) {
#ifdef __HIP_PLATFORM_AMD__
int supports_coop_launch = 0;
HIP_CHECK(
hipDeviceGetAttribute(
&supports_coop_launch,
hipDeviceAttributeCooperativeMultiDeviceLaunch,
deviceID));
if(!supports_coop_launch) {
std::cout << "Skipping, device " << deviceID << " does not support cooperative groups"
<< std::endl;
}
else
#endif
{
std::cout << deviceID << std::endl;
// Collect valid deviceIDs.
deviceIDs.push_back(deviceID);
}
}
**Kernel launch**
.. tab-set::
.. tab-item:: Thread-block
:sync: thread-block
You can access the new block representation using the original kernel launch methods.
.. code-block:: cpp
void* params[] = {&d_vector, &d_block_reduced, &d_partition_reduced};
// Launching kernel from host.
HIP_CHECK(hipLaunchKernelGGL(vector_reduce_kernel<partition_size>,
dim3(num_blocks),
dim3(threads_per_block),
0,
hipStreamDefault,
&d_vector,
&d_block_reduced,
&d_partition_reduced));
.. tab-item:: Grid
:sync: grid
Launch the cooperative kernel on a single GPU:
.. code-block:: cpp
void* params[] = {};
// Launching kernel from host.
HIP_CHECK(hipLaunchCooperativeKernel(vector_reduce_kernel<partition_size>,
dim3(num_blocks),
dim3(threads_per_block),
0,
0,
hipStreamDefault));
.. tab-item:: Multi-grid
:sync: multi-grid
Launch the cooperative kernel over multiple GPUs:
.. code-block:: cpp
hipLaunchParams *launchParamsList = (hipLaunchParams*)malloc(sizeof(hipLaunchParams) * deviceIDs.size());
for(int deviceID : deviceIDs) {
// Set device
HIP_CHECK(hipSetDevice(deviceID));
// Create stream
hipStream_t stream;
HIP_CHECK(hipStreamCreate(&stream));
// Parameters
void* params[] = {&(d_vector[deviceID]), &(d_block_reduced[deviceID]), &(d_partition_reduced[deviceID])};
// Set launchParams
launchParamsList[deviceID].func = (void*)vector_reduce_kernel<partition_size>;
launchParamsList[deviceID].gridDim = dim3(1);
launchParamsList[deviceID].blockDim = dim3(threads_per_block);
launchParamsList[deviceID].sharedMem = 0;
launchParamsList[deviceID].stream = stream;
launchParamsList[deviceID].args = params;
}
HIP_CHECK(hipLaunchCooperativeKernelMultiDevice(launchParamsList,
(int)deviceIDs.size(),
hipCooperativeLaunchMultiDeviceNoPreSync));
**Device side synchronization**
.. tab-set::
.. tab-item:: Thread-block
:sync: thread-block
The device side code of the thread_block synchronization over single GPUs:
.. code-block:: cpp
thread_block g = this_thread_block();
g.sync();
.. tab-item:: Grid
:sync: grid
The device side code of the grid synchronization over single GPUs:
.. code-block:: cpp
grid_group grid = this_grid();
grid.sync();
.. tab-item:: Multi-grid
:sync: multi-grid
The device side code of the multi-grid synchronization over multiple GPUs:
.. code-block:: cpp
multi_grid_group multi_grid = this_multi_grid();
multi_grid.sync();
Unsupported NVIDIA CUDA features
================================
HIP doesn't support the following NVIDIA CUDA optional headers:
* ``cooperative_groups/memcpy_async.h``
* ``cooperative_groups/reduce.h``
* ``cooperative_groups/scan.h``
HIP doesn't support the following CUDA class in ``cooperative_groups`` namespace:
* ``cluster_group``
HIP doesn't support the following CUDA functions/operators in ``cooperative_groups`` namespace:
* ``synchronize``
* ``memcpy_async``
* ``wait`` and ``wait_prior``
* ``barrier_arrive`` and ``barrier_wait``
* ``invoke_one`` and ``invoke_one_broadcast``
* ``reduce``
* ``reduce_update_async`` and ``reduce_store_async``
* Reduce operators ``plus`` , ``less`` , ``greater`` , ``bit_and`` , ``bit_xor`` and ``bit_or``
* ``inclusive_scan`` and ``exclusive_scan``
projects/hip/docs/how-to/faq.md
+1 -1
Melihat File
@@ -38,7 +38,7 @@ See the [API Support Table](https://github.com/ROCm/HIPIFY/blob/amd-staging/docs
* Virtual functions, indirect functions and try/catch (CUDA 4.0)
* `__prof_trigger`
* PTX assembly (CUDA 4.0). HIP-Clang supports inline GCN assembly.
* Several kernel features are under development. See the {doc}`/reference/kernel_language` for more information.
* Several kernel features are under development. See the {doc}`/reference/cpp_language_extensions` for more information.
## Is HIP a drop-in replacement for CUDA?
projects/hip/docs/how-to/hip_porting_driver_api.md
-304
Melihat File
@@ -1,304 +0,0 @@
# Porting CUDA Driver API
## Introduction to the CUDA Driver and Runtime APIs
CUDA provides a separate CUDA Driver and Runtime APIs. The two APIs have significant overlap in functionality:
* Both APIs support events, streams, memory management, memory copy, and error handling.
* Both APIs deliver similar performance.
* Driver APIs calls begin with the prefix `cu` while Runtime APIs begin with the prefix `cuda`. For example, the Driver API API contains `cuEventCreate` while the Runtime API contains `cudaEventCreate`, with similar functionality.
* The Driver API defines a different but largely overlapping error code space than the Runtime API, and uses a different coding convention. For example, Driver API defines `CUDA_ERROR_INVALID_VALUE` while the Runtime API defines `cudaErrorInvalidValue`
The Driver API offers two additional pieces of functionality not provided by the Runtime API: `cuModule` and `cuCtx` APIs.
### `cuModule` API
The Module section of the Driver API provides additional control over how and when accelerator code objects are loaded.
For example, the driver API allows code objects to be loaded from files or memory pointers.
Symbols for kernels or global data can be extracted from the loaded code objects.
In contrast, the Runtime API automatically loads and (if necessary) compiles all of the kernels from an executable binary when run.
In this mode, NVCC must be used to compile kernel code so the automatic loading can function correctly.
Both Driver and Runtime APIs define a function for launching kernels (called `cuLaunchKernel` or `cudaLaunchKernel`.
The kernel arguments and the execution configuration (grid dimensions, group dimensions, dynamic shared memory, and stream) are passed as arguments to the launch function.
The Runtime additionally provides the `<<< >>>` syntax for launching kernels, which resembles a special function call and is easier to use than explicit launch API (in particular with respect to handling of kernel arguments).
However, this syntax is not standard C++ and is available only when NVCC is used to compile the host code.
The Module features are useful in an environment which generates the code objects directly, such as a new accelerator language front-end.
Here, NVCC is not used. Instead, the environment may have a different kernel language or different compilation flow.
Other environments have many kernels and do not want them to be all loaded automatically.
The Module functions can be used to load the generated code objects and launch kernels.
As we will see below, HIP defines a Module API which provides similar explicit control over code object management.
### `cuCtx` API
The Driver API defines "Context" and "Devices" as separate entities.
Contexts contain a single device, and a device can theoretically have multiple contexts.
Each context contains a set of streams and events specific to the context.
Historically contexts also defined a unique address space for the GPU, though this may no longer be the case in Unified Memory platforms (since the CPU and all the devices in the same process share a single unified address space).
The Context APIs also provide a mechanism to switch between devices, which allowed a single CPU thread to send commands to different GPUs.
HIP as well as a recent versions of CUDA Runtime provide other mechanisms to accomplish this feat - for example using streams or `cudaSetDevice`.
The CUDA Runtime API unifies the Context API with the Device API. This simplifies the APIs and has little loss of functionality since each Context can contain a single device, and the benefits of multiple contexts has been replaced with other interfaces.
HIP provides a context API to facilitate easy porting from existing Driver codes.
In HIP, the `Ctx` functions largely provide an alternate syntax for changing the active device.
Most new applications will prefer to use `hipSetDevice` or the stream APIs , therefore HIP has marked `hipCtx` APIs as **deprecated**. Support for these APIs may not be available in future releases. For more details on deprecated APIs please refer [HIP deprecated APIs](https://github.com/ROCm/HIP/blob/develop/docs/reference/deprecated_api_list.md).
## HIP Module and `Ctx` APIs
Rather than present two separate APIs, HIP extends the HIP API with new APIs for Modules and `Ctx` control.
### `hipModule` API
Like the CUDA Driver API, the Module API provides additional control over how code is loaded, including options to load code from files or from in-memory pointers.
NVCC and HIP-Clang target different architectures and use different code object formats: NVCC is `cubin` or `ptx` files, while the HIP-Clang path is the `hsaco` format.
The external compilers which generate these code objects are responsible for generating and loading the correct code object for each platform.
Notably, there is not a fat binary format that can contain code for both NVCC and HIP-Clang platforms. The following table summarizes the formats used on each platform:
| Format | APIs | NVCC | HIP-CLANG |
| --- | --- | --- | --- |
| Code Object | `hipModuleLoad`, `hipModuleLoadData` | `.cubin` or PTX text | `.hsaco` |
| Fat Binary | `hipModuleLoadFatBin` | `.fatbin` | `.hip_fatbin` |
`hipcc` uses HIP-Clang or NVCC to compile host codes. Both of these may embed code objects into the final executable, and these code objects will be automatically loaded when the application starts.
The `hipModule` API can be used to load additional code objects, and in this way provides an extended capability to the automatically loaded code objects.
HIP-Clang allows both of these capabilities to be used together, if desired. Of course it is possible to create a program with no kernels and thus no automatic loading.
### `hipCtx` API
HIP provides a `Ctx` API as a thin layer over the existing Device functions. This `Ctx` API can be used to set the current context, or to query properties of the device associated with the context.
The current context is implicitly used by other APIs such as `hipStreamCreate`.
### hipify translation of CUDA Driver API
The HIPIFY tools convert CUDA Driver APIs for streams, events, modules, devices, memory management, context, profiler to the equivalent HIP driver calls. For example, `cuEventCreate` will be translated to `hipEventCreate`.
HIPIFY tools also convert error codes from the Driver namespace and coding convention to the equivalent HIP error code. Thus, HIP unifies the APIs for these common functions.
The memory copy API requires additional explanation. The CUDA driver includes the memory direction in the name of the API (`cuMemcpyH2D`) while the CUDA driver API provides a single memory copy API with a parameter that specifies the direction and additionally supports a "default" direction where the runtime determines the direction automatically.
HIP provides APIs with both styles: for example, `hipMemcpyH2D` as well as `hipMemcpy`.
The first flavor may be faster in some cases since they avoid host overhead to detect the different memory directions.
HIP defines a single error space, and uses camel-case for all errors (i.e. `hipErrorInvalidValue`).
#### Address Spaces
HIP-Clang defines a process-wide address space where the CPU and all devices allocate addresses from a single unified pool.
Thus addresses may be shared between contexts, and unlike the original CUDA definition a new context does not create a new address space for the device.
#### Using `hipModuleLaunchKernel`
`hipModuleLaunchKernel` is `cuLaunchKernel` in HIP world. It takes the same arguments as `cuLaunchKernel`.
#### Additional Information
* HIP-Clang creates a primary context when the HIP API is called. So in a pure driver API code, HIP-Clang will create a primary context while HIP/NVCC will have empty context stack.
HIP-Clang will push primary context to context stack when it is empty. This can have subtle differences on applications which mix the runtime and driver APIs.
### `hip-clang` Implementation Notes
#### `.hip_fatbin`
hip-clang links device code from different translation units together. For each device target, a code object is generated. Code objects for different device targets are bundled by `clang-offload-bundler` as one fatbinary, which is embeded as a global symbol `__hip_fatbin` in the `.hip_fatbin` section of the ELF file of the executable or shared object.
#### Initialization and Termination Functions
hip-clang generates initialization and termination functions for each translation unit for host code compilation. The initialization functions call `__hipRegisterFatBinary` to register the fatbinary embeded in the ELF file. They also call `__hipRegisterFunction` and `__hipRegisterVar` to register kernel functions and device side global variables. The termination functions call `__hipUnregisterFatBinary`.
hip-clang emits a global variable `__hip_gpubin_handle` of void** type with linkonce linkage and inital value 0 for each host translation unit. Each initialization function checks `__hip_gpubin_handle` and register the fatbinary only if `__hip_gpubin_handle` is 0 and saves the return value of `__hip_gpubin_handle` to `__hip_gpubin_handle`. This is to guarantee that the fatbinary is only registered once. Similar check is done in the termination functions.
#### Kernel Launching
hip-clang supports kernel launching by CUDA `<<<>>>` syntax, hipLaunchKernelGGL. The latter one is macro which expand to CUDA `<<<>>>` syntax.
When the executable or shared library is loaded by the dynamic linker, the initialization functions are called. In the initialization functions, when `__hipRegisterFatBinary` is called, the code objects containing all kernels are loaded; when `__hipRegisterFunction` is called, the stub functions are associated with the corresponding kernels in code objects.
hip-clang implements two sets of kernel launching APIs.
By default, in the host code, for the `<<<>>>` statement, hip-clang first emits call of `hipConfigureCall` to set up the threads and grids, then emits call of the stub function with the given arguments. In the stub function, `hipSetupArgument` is called for each kernel argument, then `hipLaunchByPtr` is called with a function pointer to the stub function. In `hipLaunchByPtr`, the real kernel associated with the stub function is launched.
### NVCC Implementation Notes
#### Interoperation between HIP and CUDA Driver
CUDA applications may want to mix CUDA driver code with HIP code (see example below). This table shows the type equivalence to enable this interaction.
|**HIP Type** |**CU Driver Type**|**CUDA Runtime Type**|
| ---- | ---- | ---- |
| `hipModule_t` | `CUmodule` | |
| `hipFunction_t` | `CUfunction` | |
| `hipCtx_t` | `CUcontext` | |
| `hipDevice_t` | `CUdevice` | |
| `hipStream_t` | `CUstream` | `cudaStream_t` |
| `hipEvent_t` | `CUevent` | `cudaEvent_t` |
| `hipArray` | `CUarray` | `cudaArray` |
#### Compilation Options
The `hipModule_t` interface does not support `cuModuleLoadDataEx` function, which is used to control PTX compilation options.
HIP-Clang does not use PTX and does not support these compilation options.
In fact, HIP-Clang code objects always contain fully compiled ISA and do not require additional compilation as a part of the load step.
The corresponding HIP function `hipModuleLoadDataEx` behaves as `hipModuleLoadData` on HIP-Clang path (compilation options are not used) and as `cuModuleLoadDataEx` on NVCC path.
For example (CUDA):
```cpp
CUmodule module;
void *imagePtr = ...; // Somehow populate data pointer with code object
const int numOptions = 1;
CUJit_option options[numOptions];
void * optionValues[numOptions];
options[0] = CU_JIT_MAX_REGISTERS;
unsigned maxRegs = 15;
optionValues[0] = (void*)(&maxRegs);
cuModuleLoadDataEx(module, imagePtr, numOptions, options, optionValues);
CUfunction k;
cuModuleGetFunction(&k, module, "myKernel");
```
HIP:
```cpp
hipModule_t module;
void *imagePtr = ...; // Somehow populate data pointer with code object
const int numOptions = 1;
hipJitOption options[numOptions];
void * optionValues[numOptions];
options[0] = hipJitOptionMaxRegisters;
unsigned maxRegs = 15;
optionValues[0] = (void*)(&maxRegs);
// hipModuleLoadData(module, imagePtr) will be called on HIP-Clang path, JIT options will not be used, and
// cupModuleLoadDataEx(module, imagePtr, numOptions, options, optionValues) will be called on NVCC path
hipModuleLoadDataEx(module, imagePtr, numOptions, options, optionValues);
hipFunction_t k;
hipModuleGetFunction(&k, module, "myKernel");
```
The below sample shows how to use `hipModuleGetFunction`.
```cpp
#include<hip_runtime.h>
#include<hip_runtime_api.h>
#include<iostream>
#include<fstream>
#include<vector>
#define LEN 64
#define SIZE LEN<<2
#ifdef __HIP_PLATFORM_AMD__
#define fileName "vcpy_isa.co"
#endif
#ifdef __HIP_PLATFORM_NVIDIA__
#define fileName "vcpy_isa.ptx"
#endif
#define kernel_name "hello_world"
int main(){
float *A, *B;
hipDeviceptr_t Ad, Bd;
A = new float[LEN];
B = new float[LEN];
for(uint32_t i=0;i<LEN;i++){
A[i] = i*1.0f;
B[i] = 0.0f;
std::cout<<A[i] << " "<<B[i]<<std::endl;
}
#ifdef __HIP_PLATFORM_NVIDIA__
hipInit(0);
hipDevice_t device;
hipCtx_t context;
hipDeviceGet(&device, 0);
hipCtxCreate(&context, 0, device);
#endif
hipMalloc((void**)&Ad, SIZE);
hipMalloc((void**)&Bd, SIZE);
hipMemcpyHtoD(Ad, A, SIZE);
hipMemcpyHtoD(Bd, B, SIZE);
hipModule_t Module;
hipFunction_t Function;
hipModuleLoad(&Module, fileName);
hipModuleGetFunction(&Function, Module, kernel_name);
std::vector<void*>argBuffer(2);
memcpy(&argBuffer[0], &Ad, sizeof(void*));
memcpy(&argBuffer[1], &Bd, sizeof(void*));
size_t size = argBuffer.size()*sizeof(void*);
void *config[] = {
HIP_LAUNCH_PARAM_BUFFER_POINTER, &argBuffer[0],
HIP_LAUNCH_PARAM_BUFFER_SIZE, &size,
HIP_LAUNCH_PARAM_END
};
hipModuleLaunchKernel(Function, 1, 1, 1, LEN, 1, 1, 0, 0, NULL, (void**)&config);
hipMemcpyDtoH(B, Bd, SIZE);
for(uint32_t i=0;i<LEN;i++){
std::cout<<A[i]<<" - "<<B[i]<<std::endl;
}
#ifdef __HIP_PLATFORM_NVIDIA__
hipCtxDetach(context);
#endif
return 0;
}
```
## HIP Module and Texture Driver API
HIP supports texture driver APIs however texture reference should be declared in host scope. Following code explains the use of texture reference for `__HIP_PLATFORM_AMD__` platform.
```cpp
// Code to generate code object
#include "hip/hip_runtime.h"
extern texture<float, 2, hipReadModeElementType> tex;
__global__ void tex2dKernel(hipLaunchParm lp, float* outputData,
int width,
int height)
{
int x = blockIdx.x*blockDim.x + threadIdx.x;
int y = blockIdx.y*blockDim.y + threadIdx.y;
outputData[y*width + x] = tex2D(tex, x, y);
}
```
```cpp
// Host code:
texture<float, 2, hipReadModeElementType> tex;
void myFunc ()
{
// ...
textureReference* texref;
hipModuleGetTexRef(&texref, Module1, "tex");
hipTexRefSetAddressMode(texref, 0, hipAddressModeWrap);
hipTexRefSetAddressMode(texref, 1, hipAddressModeWrap);
hipTexRefSetFilterMode(texref, hipFilterModePoint);
hipTexRefSetFlags(texref, 0);
hipTexRefSetFormat(texref, HIP_AD_FORMAT_FLOAT, 1);
hipTexRefSetArray(texref, array, HIP_TRSA_OVERRIDE_FORMAT);
// ...
}
```
projects/hip/docs/how-to/hip_porting_driver_api.rst
+537
Melihat File
@@ -0,0 +1,537 @@
.. meta::
:description: This chapter presents how to port the CUDA driver API and showcases equivalent operations in HIP.
:keywords: AMD, ROCm, HIP, CUDA, driver API
.. _porting_driver_api:
*******************************************************************************
Porting CUDA driver API
*******************************************************************************
NVIDIA provides separate CUDA driver and runtime APIs. The two APIs have significant overlap in functionality:
* Both APIs support events, streams, memory management, memory copy, and error handling.
* Both APIs deliver similar performance.
* Driver API calls begin with the prefix ``cu``, while runtime API calls begin with the prefix ``cuda``. For example, the driver API contains ``cuEventCreate``, while the runtime API contains ``cudaEventCreate``, which has similar functionality.
* The driver API defines a different, but largely overlapping, error code space than the runtime API and uses a different coding convention. For example, the driver API defines ``CUDA_ERROR_INVALID_VALUE``, while the runtime API defines ``cudaErrorInvalidValue``.
The driver API offers two additional functionalities not provided by the runtime API: ``cuModule`` and ``cuCtx`` APIs.
cuModule API
================================================================================
The Module section of the driver API provides additional control over how and
when accelerator code objects are loaded. For example, the driver API enables
code objects to load from files or memory pointers. Symbols for kernels or
global data are extracted from the loaded code objects. In contrast, the runtime
API loads automatically and, if necessary, compiles all the kernels from an
executable binary when it runs. In this mode, kernel code must be compiled using
NVCC so that automatic loading can function correctly.
The Module features are useful in an environment that generates the code objects
directly, such as a new accelerator language front end. NVCC is not used here.
Instead, the environment might have a different kernel language or compilation
flow. Other environments have many kernels and don't want all of them to be
loaded automatically. The Module functions load the generated code objects and
launch kernels. Similar to the cuModule API, HIP defines a hipModule API that
provides similar explicit control over code object management.
.. _context_driver_api:
cuCtx API
================================================================================
The driver API defines "Context" and "Devices" as separate entities.
Contexts contain a single device, and a device can theoretically have multiple contexts.
Each context contains a set of streams and events specific to the context.
Historically, contexts also defined a unique address space for the GPU. This might no longer be the case in unified memory platforms, because the CPU and all the devices in the same process share a single unified address space.
The Context APIs also provide a mechanism to switch between devices, which enables a single CPU thread to send commands to different GPUs.
HIP and recent versions of the CUDA Runtime provide other mechanisms to accomplish this feat, for example, using streams or ``cudaSetDevice``.
The CUDA runtime API unifies the Context API with the Device API. This simplifies the APIs and has little loss of functionality. This is because each context can contain a single device, and the benefits of multiple contexts have been replaced with other interfaces.
HIP provides a Context API to facilitate easy porting from existing Driver code.
In HIP, the ``Ctx`` functions largely provide an alternate syntax for changing the active device.
Most new applications preferentially use ``hipSetDevice`` or the stream APIs. Therefore, HIP has marked the ``hipCtx`` APIs as **deprecated**. Support for these APIs might not be available in future releases. For more details on deprecated APIs, see :doc:`../reference/deprecated_api_list`.
HIP module and Ctx APIs
================================================================================
Rather than present two separate APIs, HIP extends the HIP API with new APIs for
modules and ``Ctx`` control.
hipModule API
--------------------------------------------------------------------------------
Like the CUDA driver API, the Module API provides additional control over how
code is loaded, including options to load code from files or from in-memory
pointers.
NVCC and HIP-Clang target different architectures and use different code object
formats. NVCC supports ``cubin`` or ``ptx`` files, while the HIP-Clang path uses
the ``hsaco`` format.
The external compilers which generate these code objects are responsible for
generating and loading the correct code object for each platform.
Notably, there is no fat binary format that can contain code for both NVCC and
HIP-Clang platforms. The following table summarizes the formats used on each
platform:
.. list-table:: Module formats
:header-rows: 1
* - Format
- APIs
- NVCC
- HIP-CLANG
* - Code object
- ``hipModuleLoad``, ``hipModuleLoadData``
- ``.cubin`` or PTX text
- ``.hsaco``
* - Fat binary
- ``hipModuleLoadFatBin``
- ``.fatbin``
- ``.hip_fatbin``
``hipcc`` uses HIP-Clang or NVCC to compile host code. Both of these compilers can embed code objects into the final executable. These code objects are automatically loaded when the application starts.
The ``hipModule`` API can be used to load additional code objects. When used this way, it extends the capability of the automatically loaded code objects.
HIP-Clang enables both of these capabilities to be used together. Of course, it is possible to create a program with no kernels and no automatic loading.
For module API reference, visit :ref:`module_management_reference`.
hipCtx API
--------------------------------------------------------------------------------
HIP provides a ``Ctx`` API as a thin layer over the existing device functions. The ``Ctx`` API can be used to set the current context or to query properties of the device associated with the context.
The current context is implicitly used by other APIs, such as ``hipStreamCreate``.
For context reference, visit :ref:`context_management_reference`.
HIPIFY translation of CUDA driver API
================================================================================
The HIPIFY tools convert CUDA driver APIs for streams, events, modules, devices, memory management, context, and the profiler to the equivalent HIP calls. For example, ``cuEventCreate`` is translated to ``hipEventCreate``.
HIPIFY tools also convert error codes from the driver namespace and coding conventions to the equivalent HIP error code. HIP unifies the APIs for these common functions.
The memory copy API requires additional explanation. The CUDA driver includes the memory direction in the name of the API (``cuMemcpyH2D``), while the CUDA driver API provides a single memory copy API with a parameter that specifies the direction. It also supports a "default" direction where the runtime determines the direction automatically.
HIP provides APIs with both styles, for example, ``hipMemcpyH2D`` as well as ``hipMemcpy``.
The first version might be faster in some cases because it avoids any host overhead to detect the different memory directions.
HIP defines a single error space and uses camel case for all errors (i.e. ``hipErrorInvalidValue``).
For further information, visit the :doc:`hipify:index`.
Address spaces
--------------------------------------------------------------------------------
HIP-Clang defines a process-wide address space where the CPU and all devices allocate addresses from a single unified pool.
This means addresses can be shared between contexts. Unlike the original CUDA implementation, a new context does not create a new address space for the device.
Using hipModuleLaunchKernel
--------------------------------------------------------------------------------
Both CUDA driver and runtime APIs define a function for launching kernels, called ``cuLaunchKernel`` or ``cudaLaunchKernel``. The equivalent API in HIP is ``hipModuleLaunchKernel``.
The kernel arguments and the execution configuration (grid dimensions, group dimensions, dynamic shared memory, and stream) are passed as arguments to the launch function.
The runtime API additionally provides the ``<<< >>>`` syntax for launching kernels, which resembles a special function call and is easier to use than the explicit launch API, especially when handling kernel arguments.
However, this syntax is not standard C++ and is available only when NVCC is used to compile the host code.
Additional information
--------------------------------------------------------------------------------
HIP-Clang creates a primary context when the HIP API is called. So, in pure
driver API code, HIP-Clang creates a primary context while HIP/NVCC has an empty
context stack. HIP-Clang pushes the primary context to the context stack when it
is empty. This can lead to subtle differences in applications which mix the
runtime and driver APIs.
HIP-Clang implementation notes
================================================================================
.hip_fatbin
--------------------------------------------------------------------------------
HIP-Clang links device code from different translation units together. For each
device target, it generates a code object. ``clang-offload-bundler`` bundles
code objects for different device targets into one fat binary, which is embedded
as the global symbol ``__hip_fatbin`` in the ``.hip_fatbin`` section of the ELF
file of the executable or shared object.
Initialization and termination functions
--------------------------------------------------------------------------------
HIP-Clang generates initialization and termination functions for each
translation unit for host code compilation. The initialization functions call
``__hipRegisterFatBinary`` to register the fat binary embedded in the ELF file.
They also call ``__hipRegisterFunction`` and ``__hipRegisterVar`` to register
kernel functions and device-side global variables. The termination functions
call ``__hipUnregisterFatBinary``.
HIP-Clang emits a global variable ``__hip_gpubin_handle`` of type ``void**``
with ``linkonce`` linkage and an initial value of 0 for each host translation
unit. Each initialization function checks ``__hip_gpubin_handle`` and registers
the fat binary only if ``__hip_gpubin_handle`` is 0. It saves the return value
of ``__hip_gpubin_handle`` to ``__hip_gpubin_handle``. This ensures that the fat
binary is registered once. A similar check is performed in the termination
functions.
Kernel launching
--------------------------------------------------------------------------------
HIP-Clang supports kernel launching using either the CUDA ``<<<>>>`` syntax, ``hipLaunchKernel``, or ``hipLaunchKernelGGL``. The last option is a macro which expands to the CUDA ``<<<>>>`` syntax by default. It can also be turned into a template by defining ``HIP_TEMPLATE_KERNEL_LAUNCH``.
When the executable or shared library is loaded by the dynamic linker, the initialization functions are called. In the initialization functions, the code objects containing all kernels are loaded when ``__hipRegisterFatBinary`` is called. When ``__hipRegisterFunction`` is called, the stub functions are associated with the corresponding kernels in the code objects.
HIP-Clang implements two sets of APIs for launching kernels.
By default, when HIP-Clang encounters the ``<<<>>>`` statement in the host code, it first calls ``hipConfigureCall`` to set up the threads and grids. It then calls the stub function with the given arguments. The stub function calls ``hipSetupArgument`` for each kernel argument, then calls ``hipLaunchByPtr`` with a function pointer to the stub function. In ``hipLaunchByPtr``, the actual kernel associated with the stub function is launched.
NVCC implementation notes
================================================================================
Interoperation between HIP and CUDA driver
--------------------------------------------------------------------------------
CUDA applications might want to mix CUDA driver code with HIP code (see the example below). This table shows the equivalence between CUDA and HIP types required to implement this interaction.
.. list-table:: Equivalence table between HIP and CUDA types
:header-rows: 1
* - HIP type
- CU Driver type
- CUDA Runtime type
* - ``hipModule_t``
- ``CUmodule``
-
* - ``hipFunction_t``
- ``CUfunction``
-
* - ``hipCtx_t``
- ``CUcontext``
-
* - ``hipDevice_t``
- ``CUdevice``
-
* - ``hipStream_t``
- ``CUstream``
- ``cudaStream_t``
* - ``hipEvent_t``
- ``CUevent``
- ``cudaEvent_t``
* - ``hipArray``
- ``CUarray``
- ``cudaArray``
Compilation options
--------------------------------------------------------------------------------
The ``hipModule_t`` interface does not support the ``cuModuleLoadDataEx`` function, which is used to control PTX compilation options.
HIP-Clang does not use PTX, so it does not support these compilation options.
In fact, HIP-Clang code objects contain fully compiled code for a device-specific instruction set and don't require additional compilation as a part of the load step.
The corresponding HIP function ``hipModuleLoadDataEx`` behaves like ``hipModuleLoadData`` on the HIP-Clang path (where compilation options are not used) and like ``cuModuleLoadDataEx`` on the NVCC path.
For example:
.. tab-set::
.. tab-item:: HIP
.. code-block:: cpp
hipModule_t module;
void *imagePtr = ...; // Somehow populate data pointer with code object
const int numOptions = 1;
hipJitOption options[numOptions];
void *optionValues[numOptions];
options[0] = hipJitOptionMaxRegisters;
unsigned maxRegs = 15;
optionValues[0] = (void *)(&maxRegs);
// hipModuleLoadData(module, imagePtr) will be called on HIP-Clang path, JIT
// options will not be used, and cupModuleLoadDataEx(module, imagePtr,
// numOptions, options, optionValues) will be called on NVCC path
hipModuleLoadDataEx(module, imagePtr, numOptions, options, optionValues);
hipFunction_t k;
hipModuleGetFunction(&k, module, "myKernel");
.. tab-item:: CUDA
.. code-block:: cpp
CUmodule module;
void *imagePtr = ...; // Somehow populate data pointer with code object
const int numOptions = 1;
CUJit_option options[numOptions];
void *optionValues[numOptions];
options[0] = CU_JIT_MAX_REGISTERS;
unsigned maxRegs = 15;
optionValues[0] = (void *)(&maxRegs);
cuModuleLoadDataEx(module, imagePtr, numOptions, options, optionValues);
CUfunction k;
cuModuleGetFunction(&k, module, "myKernel");
The sample below shows how to use ``hipModuleGetFunction``.
.. code-block:: cpp
#include <hip/hip_runtime.h>
#include <hip/hip_runtime_api.h>
#include <vector>
int main() {
size_t elements = 64*1024;
size_t size_bytes = elements * sizeof(float);
std::vector<float> A(elements), B(elements);
// On NVIDIA platforms the driver runtime needs to be initiated
#ifdef __HIP_PLATFORM_NVIDIA__
hipInit(0);
hipDevice_t device;
hipCtx_t context;
HIPCHECK(hipDeviceGet(&device, 0));
HIPCHECK(hipCtxCreate(&context, 0, device));
#endif
// Allocate device memory
hipDeviceptr_t d_A, d_B;
HIPCHECK(hipMalloc(&d_A, size_bytes));
HIPCHECK(hipMalloc(&d_B, size_bytes));
// Copy data to device
HIPCHECK(hipMemcpyHtoD(d_A, A.data(), size_bytes));
HIPCHECK(hipMemcpyHtoD(d_B, B.data(), size_bytes));
// Load module
hipModule_t Module;
// For AMD the module file has to contain architecture specific object codee
// For NVIDIA the module file has to contain PTX, found in e.g. "vcpy_isa.ptx"
HIPCHECK(hipModuleLoad(&Module, "vcpy_isa.co"));
// Get kernel function from the module via its name
hipFunction_t Function;
HIPCHECK(hipModuleGetFunction(&Function, Module, "hello_world"));
// Create buffer for kernel arguments
std::vector<void*> argBuffer{&d_A, &d_B};
size_t arg_size_bytes = argBuffer.size() * sizeof(void*);
// Create configuration passed to the kernel as arguments
void* config[] = {HIP_LAUNCH_PARAM_BUFFER_POINTER, argBuffer.data(),
HIP_LAUNCH_PARAM_BUFFER_SIZE, &arg_size_bytes, HIP_LAUNCH_PARAM_END};
int threads_per_block = 128;
int blocks = (elements + threads_per_block - 1) / threads_per_block;
// Actually launch kernel
HIPCHECK(hipModuleLaunchKernel(Function, blocks, 1, 1, threads_per_block, 1, 1, 0, 0, NULL, config));
HIPCHECK(hipMemcpyDtoH(A.data(), d_A, elements));
HIPCHECK(hipMemcpyDtoH(B.data(), d_B, elements));
#ifdef __HIP_PLATFORM_NVIDIA__
HIPCHECK(hipCtxDetach(context));
#endif
HIPCHECK(hipFree(d_A));
HIPCHECK(hipFree(d_B));
return 0;
}
HIP module and texture Driver API
================================================================================
HIP supports texture driver APIs. However, texture references must be declared
within the host scope. The following code demonstrates the use of texture
references for the ``__HIP_PLATFORM_AMD__`` platform.
.. code-block:: cpp
// Code to generate code object
#include "hip/hip_runtime.h"
extern texture<float, 2, hipReadModeElementType> tex;
__global__ void tex2dKernel(hipLaunchParm lp, float *outputData, int width,
int height) {
int x = blockIdx.x * blockDim.x + threadIdx.x;
int y = blockIdx.y * blockDim.y + threadIdx.y;
outputData[y * width + x] = tex2D(tex, x, y);
}
.. code-block:: cpp
// Host code:
texture<float, 2, hipReadModeElementType> tex;
void myFunc ()
{
// ...
textureReference* texref;
hipModuleGetTexRef(&texref, Module1, "tex");
hipTexRefSetAddressMode(texref, 0, hipAddressModeWrap);
hipTexRefSetAddressMode(texref, 1, hipAddressModeWrap);
hipTexRefSetFilterMode(texref, hipFilterModePoint);
hipTexRefSetFlags(texref, 0);
hipTexRefSetFormat(texref, HIP_AD_FORMAT_FLOAT, 1);
hipTexRefSetArray(texref, array, HIP_TRSA_OVERRIDE_FORMAT);
// ...
}
Driver entry point access
================================================================================
Starting from HIP version 6.2.0, support for Driver Entry Point Access is
available when using CUDA 12.0 or newer. This feature allows developers to
directly interact with the CUDA driver API, providing more control over GPU
operations.
Driver Entry Point Access provides several features:
* Retrieving the address of a runtime function
* Requesting the default stream version on a per-thread basis
* Accessing new HIP features on older toolkits with a newer driver
For driver entry point access reference, visit :cpp:func:`hipGetProcAddress`.
Address retrieval
--------------------------------------------------------------------------------
The :cpp:func:`hipGetProcAddress` function can be used to obtain the address of
a runtime function. This is demonstrated in the following example:
.. code-block:: cpp
#include <hip/hip_runtime.h>
#include <hip/hip_runtime_api.h>
#include <iostream>
typedef hipError_t (*hipInit_t)(unsigned int);
int main() {
// Initialize the HIP runtime
hipError_t res = hipInit(0);
if (res != hipSuccess) {
std::cerr << "Failed to initialize HIP runtime." << std::endl;
return 1;
}
// Get the address of the hipInit function
hipInit_t hipInitFunc;
int hipVersion = HIP_VERSION; // Use the HIP version defined in hip_runtime_api.h
uint64_t flags = 0; // No special flags
hipDriverProcAddressQueryResult symbolStatus;
res = hipGetProcAddress("hipInit", (void**)&hipInitFunc, hipVersion, flags, &symbolStatus);
if (res != hipSuccess) {
std::cerr << "Failed to get address of hipInit()." << std::endl;
return 1;
}
// Call the hipInit function using the obtained address
res = hipInitFunc(0);
if (res == hipSuccess) {
std::cout << "HIP runtime initialized successfully using hipGetProcAddress()." << std::endl;
} else {
std::cerr << "Failed to initialize HIP runtime using hipGetProcAddress()." << std::endl;
}
return 0;
}
Per-thread default stream version request
================================================================================
HIP offers functionality similar to CUDA for managing streams on a per-thread
basis. By using ``hipStreamPerThread``, each thread can independently manage its
default stream, simplifying operations. The following example demonstrates how
this feature enhances performance by reducing contention and improving
efficiency.
.. code-block:: cpp
#include <hip/hip_runtime.h>
#include <iostream>
int main() {
// Initialize the HIP runtime
hipError_t res = hipInit(0);
if (res != hipSuccess) {
std::cerr << "Failed to initialize HIP runtime." << std::endl;
return 1;
}
// Get the per-thread default stream
hipStream_t stream = hipStreamPerThread;
// Use the stream for some operation
// For example, allocate memory on the device
void* d_ptr;
size_t size = 1024;
res = hipMalloc(&d_ptr, size);
if (res != hipSuccess) {
std::cerr << "Failed to allocate memory." << std::endl;
return 1;
}
// Perform some operation using the stream
// For example, set memory on the device
res = hipMemsetAsync(d_ptr, 0, size, stream);
if (res != hipSuccess) {
std::cerr << "Failed to set memory." << std::endl;
return 1;
}
// Synchronize the stream
res = hipStreamSynchronize(stream);
if (res != hipSuccess) {
std::cerr << "Failed to synchronize stream." << std::endl;
return 1;
}
std::cout << "Operation completed successfully using per-thread default stream." << std::endl;
// Free the allocated memory
hipFree(d_ptr);
return 0;
}
Accessing new HIP features with a newer driver
================================================================================
HIP is designed to be forward compatible, allowing newer features to be utilized
with older toolkits, provided a compatible driver is present. Feature support
can be verified through runtime API functions and version checks. This approach
ensures that applications can benefit from new features and improvements in the
HIP runtime without needing to be recompiled with a newer toolkit. The function
:cpp:func:`hipGetProcAddress` enables dynamic querying and the use of newer
functions offered by the HIP runtime, even if the application was built with an
older toolkit.
An example is provided for a hypothetical ``foo()`` function.
.. code-block:: cpp
// Get the address of the foo function
foo_t fooFunc;
int hipVersion = 60300000; // Use an own HIP version number (e.g. 6.3.0)
uint64_t flags = 0; // No special flags
hipDriverProcAddressQueryResult symbolStatus;
res = hipGetProcAddress("foo", (void**)&fooFunc, hipVersion, flags, &symbolStatus);
The HIP version number is defined as an integer:
.. code-block:: cpp
HIP_VERSION=HIP_VERSION_MAJOR * 10000000 + HIP_VERSION_MINOR * 100000 + HIP_VERSION_PATCH
projects/hip/docs/how-to/hip_porting_guide.md
+3 -3
Melihat File
@@ -1,4 +1,4 @@
# HIP Porting Guide
# HIP porting guide
In addition to providing a portable C++ programming environment for GPUs, HIP is designed to ease
the porting of existing CUDA code into the HIP environment. This section describes the available tools
@@ -366,7 +366,7 @@ run hipcc when appropriate.
### ``warpSize``
Code should not assume a warp size of 32 or 64. See [Warp Cross-Lane Functions](https://rocm.docs.amd.com/projects/HIP/en/latest/reference/kernel_language.html#warp-cross-lane-functions) for information on how to write portable wave-aware code.
Code should not assume a warp size of 32 or 64. See [Warp Cross-Lane Functions](https://rocm.docs.amd.com/projects/HIP/en/latest/reference/cpp_language_extensions.html#warp-cross-lane-functions) for information on how to write portable wave-aware code.
### Kernel launch with group size > 256
@@ -403,7 +403,7 @@ Device Code:
__constant__ int Value[LEN];
__global__ void Get(hipLaunchParm lp, int *Ad)
__global__ void Get(int *Ad)
{
int tid = threadIdx.x + blockIdx.x * blockDim.x;
Ad[tid] = Value[tid];
projects/hip/docs/how-to/hip_rtc.md
+2 -2
Melihat File
@@ -1,4 +1,4 @@
# Programming for HIP Runtime Compiler (RTC)
# Programming for HIP runtime compiler (RTC)
HIP lets you compile kernels at runtime with the `hiprtc*` APIs.
Kernels can be stored as a text string and can be passed to HIPRTC APIs alongside options to guide the compilation.
@@ -6,7 +6,7 @@ Kernels can be stored as a text string and can be passed to HIPRTC APIs alongsid
NOTE:
* This library can be used on systems without HIP installed nor AMD GPU driver installed at all (offline compilation). Therefore, it does not depend on any HIP runtime library.
* But it does depend on COMGr. You may try to statically link COMGr into HIPRTC to avoid any ambiguity.
* But it does depend on Code Object Manager (comgr). You may try to statically link comgr into HIPRTC to avoid any ambiguity.
* Developers can decide to bundle this library with their application.
## Compilation APIs
projects/hip/docs/how-to/hipgraph.rst
+516
Melihat File
@@ -0,0 +1,516 @@
.. meta::
:description: This chapter describes how to use HIP graphs and highlights their use cases.
:keywords: ROCm, HIP, graph, stream
.. _how_to_HIP_graph:
********************************************************************************
HIP graphs
********************************************************************************
.. note::
The HIP graph API is currently in Beta. Some features can change and might
have outstanding issues. Not all features supported by CUDA graphs are yet
supported. For a list of all currently supported functions see the
:doc:`HIP graph API documentation<../doxygen/html/group___graph>`.
HIP graphs are an alternative way of executing tasks on a GPU that can provide
performance benefits over launching kernels using the standard
method via streams. A HIP graph is made up of nodes and edges. The nodes of a
HIP graph represent the operations performed, while the edges mark dependencies
between those operations.
The nodes can be one of the following:
- empty nodes
- nested graphs
- kernel launches
- host-side function calls
- HIP memory functions (copy, memset, ...)
- HIP events
- signalling or waiting on external semaphores
.. note::
The available node types are specified by :cpp:enum:`hipGraphNodeType`.
The following figure visualizes the concept of graphs, compared to using streams.
.. figure:: ../data/how-to/hipgraph/hip_graph.svg
:alt: Diagram depicting the difference between using streams to execute
kernels with dependencies, resolved by explicitly synchronizing,
or using graphs, where the edges denote the dependencies.
The standard method of launching kernels incurs a small overhead for each
iteration of the operation involved. That overhead is negligible, when the
kernel is launched directly with the HIP C/C++ API, but depending on the
framework used, there can be several levels of redirection, until the actual
kernel is launched by the HIP runtime, leading to significant overhead.
Especially for some AI frameworks, a GPU kernel might run faster than the time
it takes for the framework to set up and launch the kernel, and so the overhead
of repeatedly launching kernels can have a significant impact on performance.
HIP graphs are designed to address this issue, by predefining the HIP API calls
and their dependencies with a graph, and performing most of the initialization
beforehand. Launching a graph only requires a single call, after which the
HIP runtime takes care of executing the operations within the graph.
Graphs can provide additional performance benefits, by enabling optimizations
that are only possible when knowing the dependencies between the operations.
.. figure:: ../data/how-to/hipgraph/hip_graph_speedup.svg
:alt: Diagram depicting the speed up achievable with HIP graphs compared to
HIP streams when launching many short-running kernels.
Qualitative presentation of the execution time of many short-running kernels
when launched using HIP stream versus HIP graph. This does not include the
time needed to set up the graph.
Using HIP graphs
================================================================================
There are two different ways of creating graphs: Capturing kernel launches from
a stream, or explicitly creating graphs. The difference between the two
approaches is explained later in this chapter.
The general flow for using HIP graphs includes the following steps.
#. Create a :cpp:type:`hipGraph_t` graph template using one of the two approaches described in this chapter
#. Create a :cpp:type:`hipGraphExec_t` executable instance of the graph template using :cpp:func:`hipGraphInstantiate`
#. Use :cpp:func:`hipGraphLaunch` to launch the executable graph to a stream
#. After execution completes free and destroy graph resources
The first two steps are the initial setup and only need to be executed once. First
step is the definition of the operations (nodes) and the dependencies (edges)
between them. The second step is the instantiation of the graph. This takes care
of validating and initializing the graph, to reduce the overhead when executing
the graph. The third step is the execution of the graph, which takes care of
launching all the kernels and executing the operations while respecting their
dependencies and necessary synchronizations as specified.
Because HIP graphs require some setup and initialization overhead before their
first execution, graphs only provide a benefit for workloads that require
many iterations to complete.
In both methods the :cpp:type:`hipGraph_t` template for a graph is used to define the graph.
In order to actually launch a graph, the template needs to be instantiated using
:cpp:func:`hipGraphInstantiate`, which results in an executable graph of type :cpp:type:`hipGraphExec_t`.
This executable graph can then be launched with :cpp:func:`hipGraphLaunch`, replaying the
operations within the graph. Note, that launching graphs is fundamentally no
different to executing other HIP functions on a stream, except for the fact,
that scheduling the operations within the graph encompasses less overhead and
can enable some optimizations, but they still need to be associated with a stream for execution.
Memory management
--------------------------------------------------------------------------------
Memory that is used by operations in graphs can either be pre-allocated or
managed within the graph. Graphs can contain nodes that take care of allocating
memory on the device or copying memory between the host and the device.
Whether you want to pre-allocate the memory or manage it within the graph
depends on the use-case. If the graph is executed in a tight loop the
performance is usually better when the memory is preallocated, so that it
does not need to be reallocated in every iteration.
The same rules as for normal memory allocations apply for memory allocated and
freed by nodes, meaning that the nodes that access memory allocated in a graph
must be ordered after allocation and before freeing.
Memory management within the graph enables the runtime to take care of memory reuse and optimizations.
The lifetime of memory managed in a graph begins when the execution reaches the
node allocating the memory, and ends when either reaching the corresponding
free node within the graph, or after graph execution when a corresponding
:cpp:func:`hipFreeAsync` or :cpp:func:`hipFree` call is reached.
The memory can also be freed with a free node in a different graph that is
associated with the same memory address.
Unlike device memory that is not associated with a graph, this does not necessarily
mean that the freed memory is returned back to the operating system immediately.
Graphs can retain a memory pool for quickly reusing memory within the graph.
This can be especially useful when memory is freed and reallocated later on
within a graph, as that memory doesn't have to be requested from the operating system.
It also potentially reduces the total memory footprint of the graph, by reusing the same memory.
The amount of memory allocated for graph memory pools on a specific device can
be queried using :cpp:func:`hipDeviceGetGraphMemAttribute`.
In order to return the freed memory :cpp:func:`hipDeviceGraphMemTrim` can be used.
This will return any memory that is not in active use by graphs.
These memory allocations can also be set up to allow access from multiple GPUs,
just like normal allocations. HIP then takes care of allocating and mapping the
memory to the GPUs. When capturing a graph from a stream, the node sets the
accessibility according to :cpp:func:`hipMemPoolSetAccess` at the time of capturing.
Capture graphs from a stream
================================================================================
The easy way to integrate HIP graphs into already existing code is to use
:cpp:func:`hipStreamBeginCapture` and :cpp:func:`hipStreamEndCapture` to obtain a :cpp:type:`hipGraph_t`
graph template that includes the captured operations.
When starting to capture operations for a graph using :cpp:func:`hipStreamBeginCapture`,
the operations assigned to the stream are captured into a graph instead of being
executed. The associated graph is returned when calling :cpp:func:`hipStreamEndCapture`, which
also stops capturing operations.
In order to capture to an already existing graph use :cpp:func:`hipStreamBeginCaptureToGraph`.
The functions assigned to the capturing stream are not executed, but instead are
captured and defined as nodes in the graph, to be run when the instantiated
graph is launched.
Functions must be associated with a stream in order to be captured.
This means that non-HIP API functions are not captured by default, but are
executed as standard functions when encountered and not added to the graph.
In order to assign host functions to a stream use
:cpp:func:`hipLaunchHostFunc`, as shown in the following code example.
They will then be captured and defined as a host node in the resulting graph,
and won't be executed when encountered.
Synchronous HIP API calls that are implicitly assigned to the default stream are
not permitted while capturing a stream and will return an error. This is
because they implicitly synchronize and cause a dependency that can not be
captured within the stream. This includes functions like :cpp:func:`hipMalloc`,
:cpp:func:`hipMemcpy` and :cpp:func:`hipFree`. In order to capture these to the stream, replace
them with the corresponding asynchronous calls like :cpp:func:`hipMallocAsync`, :cpp:func:`hipMemcpyAsync` or :cpp:func:`hipFreeAsync`.
The general flow for using stream capture to create a graph template is:
#. Create a stream from which to capture the operations
#. Call :cpp:func:`hipStreamBeginCapture` before the first operation to be captured
#. Call :cpp:func:`hipStreamEndCapture` after the last operation to be captured
#. Define a :cpp:type:`hipGraph_t` graph template to which :cpp:func:`hipStreamEndCapture`
passes the captured graph
The following code is an example of how to use the HIP graph API to capture a
graph from a stream.
.. code-block:: cpp
#include <hip/hip_runtime.h>
#include <vector>
#include <iostream>
#define HIP_CHECK(expression) \
{ \
const hipError_t status = expression; \
if(status != hipSuccess){ \
std::cerr << "HIP error " \
<< status << ": " \
<< hipGetErrorString(status) \
<< " at " << __FILE__ << ":" \
<< __LINE__ << std::endl; \
} \
}
__global__ void kernelA(double* arrayA, size_t size){
const size_t x = threadIdx.x + blockDim.x * blockIdx.x;
if(x < size){arrayA[x] *= 2.0;}
};
__global__ void kernelB(int* arrayB, size_t size){
const size_t x = threadIdx.x + blockDim.x * blockIdx.x;
if(x < size){arrayB[x] = 3;}
};
__global__ void kernelC(double* arrayA, const int* arrayB, size_t size){
const size_t x = threadIdx.x + blockDim.x * blockIdx.x;
if(x < size){arrayA[x] += arrayB[x];}
};
struct set_vector_args{
std::vector<double>& h_array;
double value;
};
void set_vector(void* args){
set_vector_args h_args{*(reinterpret_cast<set_vector_args*>(args))};
std::vector<double>& vec{h_args.h_array};
vec.assign(vec.size(), h_args.value);
}
int main(){
constexpr int numOfBlocks = 1024;
constexpr int threadsPerBlock = 1024;
constexpr size_t arraySize = 1U << 20;
// This example assumes that kernelA operates on data that needs to be initialized on
// and copied from the host, while kernelB initializes the array that is passed to it.
// Both arrays are then used as input to kernelC, where arrayA is also used as
// output, that is copied back to the host, while arrayB is only read from and not modified.
double* d_arrayA;
int* d_arrayB;
std::vector<double> h_array(arraySize);
constexpr double initValue = 2.0;
hipStream_t captureStream;
HIP_CHECK(hipStreamCreate(&captureStream));
// Start capturing the operations assigned to the stream
HIP_CHECK(hipStreamBeginCapture(captureStream, hipStreamCaptureModeGlobal));
// hipMallocAsync and hipMemcpyAsync are needed, to be able to assign it to a stream
HIP_CHECK(hipMallocAsync(&d_arrayA, arraySize*sizeof(double), captureStream));
HIP_CHECK(hipMallocAsync(&d_arrayB, arraySize*sizeof(int), captureStream));
// Assign host function to the stream
// Needs a custom struct to pass the arguments
set_vector_args args{h_array, initValue};
HIP_CHECK(hipLaunchHostFunc(captureStream, set_vector, &args));
HIP_CHECK(hipMemcpyAsync(d_arrayA, h_array.data(), arraySize*sizeof(double), hipMemcpyHostToDevice, captureStream));
kernelA<<<numOfBlocks, threadsPerBlock, 0, captureStream>>>(d_arrayA, arraySize);
kernelB<<<numOfBlocks, threadsPerBlock, 0, captureStream>>>(d_arrayB, arraySize);
kernelC<<<numOfBlocks, threadsPerBlock, 0, captureStream>>>(d_arrayA, d_arrayB, arraySize);
HIP_CHECK(hipMemcpyAsync(h_array.data(), d_arrayA, arraySize*sizeof(*d_arrayA), hipMemcpyDeviceToHost, captureStream));
HIP_CHECK(hipFreeAsync(d_arrayA, captureStream));
HIP_CHECK(hipFreeAsync(d_arrayB, captureStream));
// Stop capturing
hipGraph_t graph;
HIP_CHECK(hipStreamEndCapture(captureStream, &graph));
// Create an executable graph from the captured graph
hipGraphExec_t graphExec;
HIP_CHECK(hipGraphInstantiate(&graphExec, graph, nullptr, nullptr, 0));
// The graph template can be deleted after the instantiation if it's not needed for later use
HIP_CHECK(hipGraphDestroy(graph));
// Actually launch the graph. The stream does not have
// to be the same as the one used for capturing.
HIP_CHECK(hipGraphLaunch(graphExec, captureStream));
// Verify results
constexpr double expected = initValue * 2.0 + 3;
bool passed = true;
for(size_t i = 0; i < arraySize; ++i){
if(h_array[i] != expected){
passed = false;
std::cerr << "Validation failed! Expected " << expected << " got " << h_array[0] << std::endl;
break;
}
}
if(passed){
std::cerr << "Validation passed." << std::endl;
}
// Free graph and stream resources after usage
HIP_CHECK(hipGraphExecDestroy(graphExec));
HIP_CHECK(hipStreamDestroy(captureStream));
}
Explicit graph creation
================================================================================
Graphs can also be created directly using the HIP graph API, giving more
fine-grained control over the graph. In this case, the graph nodes are created
explicitly, together with their parameters and dependencies, which specify the
edges of the graph, thereby forming the graph structure.
The nodes are represented by the generic :cpp:type:`hipGraphNode_t` type. The actual
node type is implicitly defined by the specific function used to add the node to
the graph, for example :cpp:func:`hipGraphAddKernelNode` See the
:doc:`HIP graph API documentation<../doxygen/html/group___graph>` for the
available functions, they are of type ``hipGraphAdd{Type}Node``. Each type of
node also has a predefined set of parameters depending on the operation, for
example :cpp:class:`hipKernelNodeParams` for a kernel launch. See the
:doc:`documentation for the general hipGraphNodeParams type<../doxygen/html/structhip_graph_node_params>`
for a list of available parameter types and their members.
The general flow for explicitly creating a graph is usually:
#. Create a graph :cpp:type:`hipGraph_t`
#. Create the nodes and their parameters and add them to the graph
#. Define a :cpp:type:`hipGraphNode_t`
#. Define the parameter struct for the desired operation, by explicitly setting the appropriate struct's members.
#. Use the appropriate ``hipGraphAdd{Type}Node`` function to add the node to the graph.
#. The dependencies can be defined when adding the node to the graph, or afterwards by using :cpp:func:`hipGraphAddDependencies`
The following code example demonstrates how to explicitly create nodes in order to create a graph.
.. code-block:: cpp
#include <hip/hip_runtime.h>
#include <vector>
#include <iostream>
#define HIP_CHECK(expression) \
{ \
const hipError_t status = expression; \
if(status != hipSuccess){ \
std::cerr << "HIP error " \
<< status << ": " \
<< hipGetErrorString(status) \
<< " at " << __FILE__ << ":" \
<< __LINE__ << std::endl; \
} \
}
__global__ void kernelA(double* arrayA, size_t size){
const size_t x = threadIdx.x + blockDim.x * blockIdx.x;
if(x < size){arrayA[x] *= 2.0;}
};
__global__ void kernelB(int* arrayB, size_t size){
const size_t x = threadIdx.x + blockDim.x * blockIdx.x;
if(x < size){arrayB[x] = 3;}
};
__global__ void kernelC(double* arrayA, const int* arrayB, size_t size){
const size_t x = threadIdx.x + blockDim.x * blockIdx.x;
if(x < size){arrayA[x] += arrayB[x];}
};
struct set_vector_args{
std::vector<double>& h_array;
double value;
};
void set_vector(void* args){
set_vector_args h_args{*(reinterpret_cast<set_vector_args*>(args))};
std::vector<double>& vec{h_args.h_array};
vec.assign(vec.size(), h_args.value);
}
int main(){
constexpr int numOfBlocks = 1024;
constexpr int threadsPerBlock = 1024;
size_t arraySize = 1U << 20;
// The pointers to the device memory don't need to be declared here,
// they are contained within the hipMemAllocNodeParams as the dptr member
std::vector<double> h_array(arraySize);
constexpr double initValue = 2.0;
// Create graph an empty graph
hipGraph_t graph;
HIP_CHECK(hipGraphCreate(&graph, 0));
// Parameters to allocate arrays
hipMemAllocNodeParams allocArrayAParams{};
allocArrayAParams.poolProps.allocType = hipMemAllocationTypePinned;
allocArrayAParams.poolProps.location.type = hipMemLocationTypeDevice;
allocArrayAParams.poolProps.location.id = 0; // GPU on which memory resides
allocArrayAParams.bytesize = arraySize * sizeof(double);
hipMemAllocNodeParams allocArrayBParams{};
allocArrayBParams.poolProps.allocType = hipMemAllocationTypePinned;
allocArrayBParams.poolProps.location.type = hipMemLocationTypeDevice;
allocArrayBParams.poolProps.location.id = 0; // GPU on which memory resides
allocArrayBParams.bytesize = arraySize * sizeof(int);
// Add the allocation nodes to the graph. They don't have any dependencies
hipGraphNode_t allocNodeA, allocNodeB;
HIP_CHECK(hipGraphAddMemAllocNode(&allocNodeA, graph, nullptr, 0, &allocArrayAParams));
HIP_CHECK(hipGraphAddMemAllocNode(&allocNodeB, graph, nullptr, 0, &allocArrayBParams));
// Parameters for the host function
// Needs custom struct to pass the arguments
set_vector_args args{h_array, initValue};
hipHostNodeParams hostParams{};
hostParams.fn = set_vector;
hostParams.userData = static_cast<void*>(&args);
// Add the host node that initializes the host array. It also doesn't have any dependencies
hipGraphNode_t hostNode;
HIP_CHECK(hipGraphAddHostNode(&hostNode, graph, nullptr, 0, &hostParams));
// Add memory copy node, that copies the initialized host array to the device.
// It has to wait for the host array to be initialized and the device memory to be allocated
hipGraphNode_t cpyNodeDependencies[] = {allocNodeA, hostNode};
hipGraphNode_t cpyToDevNode;
HIP_CHECK(hipGraphAddMemcpyNode1D(&cpyToDevNode, graph, cpyNodeDependencies, 1, allocArrayAParams.dptr, h_array.data(), arraySize * sizeof(double), hipMemcpyHostToDevice));
// Parameters for kernelA
hipKernelNodeParams kernelAParams;
void* kernelAArgs[] = {&allocArrayAParams.dptr, static_cast<void*>(&arraySize)};
kernelAParams.func = reinterpret_cast<void*>(kernelA);
kernelAParams.gridDim = numOfBlocks;
kernelAParams.blockDim = threadsPerBlock;
kernelAParams.sharedMemBytes = 0;
kernelAParams.kernelParams = kernelAArgs;
kernelAParams.extra = nullptr;
// Add the node for kernelA. It has to wait for the memory copy to finish, as it depends on the values from the host array.
hipGraphNode_t kernelANode;
HIP_CHECK(hipGraphAddKernelNode(&kernelANode, graph, &cpyToDevNode, 1, &kernelAParams));
// Parameters for kernelB
hipKernelNodeParams kernelBParams;
void* kernelBArgs[] = {&allocArrayBParams.dptr, static_cast<void*>(&arraySize)};
kernelBParams.func = reinterpret_cast<void*>(kernelB);
kernelBParams.gridDim = numOfBlocks;
kernelBParams.blockDim = threadsPerBlock;
kernelBParams.sharedMemBytes = 0;
kernelBParams.kernelParams = kernelBArgs;
kernelBParams.extra = nullptr;
// Add the node for kernelB. It only has to wait for the memory to be allocated, as it initializes the array.
hipGraphNode_t kernelBNode;
HIP_CHECK(hipGraphAddKernelNode(&kernelBNode, graph, &allocNodeB, 1, &kernelBParams));
// Parameters for kernelC
hipKernelNodeParams kernelCParams;
void* kernelCArgs[] = {&allocArrayAParams.dptr, &allocArrayBParams.dptr, static_cast<void*>(&arraySize)};
kernelCParams.func = reinterpret_cast<void*>(kernelC);
kernelCParams.gridDim = numOfBlocks;
kernelCParams.blockDim = threadsPerBlock;
kernelCParams.sharedMemBytes = 0;
kernelCParams.kernelParams = kernelCArgs;
kernelCParams.extra = nullptr;
// Add the node for kernelC. It has to wait on both kernelA and kernelB to finish, as it depends on their results.
hipGraphNode_t kernelCNode;
hipGraphNode_t kernelCDependencies[] = {kernelANode, kernelBNode};
HIP_CHECK(hipGraphAddKernelNode(&kernelCNode, graph, kernelCDependencies, 1, &kernelCParams));
// Copy the results back to the host. Has to wait for kernelC to finish.
hipGraphNode_t cpyToHostNode;
HIP_CHECK(hipGraphAddMemcpyNode1D(&cpyToHostNode, graph, &kernelCNode, 1, h_array.data(), allocArrayAParams.dptr, arraySize * sizeof(double), hipMemcpyDeviceToHost));
// Free array of allocNodeA. It needs to wait for the copy to finish, as kernelC stores its results in it.
hipGraphNode_t freeNodeA;
HIP_CHECK(hipGraphAddMemFreeNode(&freeNodeA, graph, &cpyToHostNode, 1, allocArrayAParams.dptr));
// Free array of allocNodeB. It only needs to wait for kernelC to finish, as it is not written back to the host.
hipGraphNode_t freeNodeB;
HIP_CHECK(hipGraphAddMemFreeNode(&freeNodeB, graph, &kernelCNode, 1, allocArrayBParams.dptr));
// Instantiate the graph in order to execute it
hipGraphExec_t graphExec;
HIP_CHECK(hipGraphInstantiate(&graphExec, graph, nullptr, nullptr, 0));
// The graph can be freed after the instantiation if it's not needed for other purposes
HIP_CHECK(hipGraphDestroy(graph));
// Actually launch the graph
hipStream_t graphStream;
HIP_CHECK(hipStreamCreate(&graphStream));
HIP_CHECK(hipGraphLaunch(graphExec, graphStream));
// Verify results
constexpr double expected = initValue * 2.0 + 3;
bool passed = true;
for(size_t i = 0; i < arraySize; ++i){
if(h_array[i] != expected){
passed = false;
std::cerr << "Validation failed! Expected " << expected << " got " << h_array[0] << std::endl;
break;
}
}
if(passed){
std::cerr << "Validation passed." << std::endl;
}
HIP_CHECK(hipGraphExecDestroy(graphExec));
HIP_CHECK(hipStreamDestroy(graphStream));
}
projects/hip/docs/how-to/performance_guidelines.rst
+213 -199
Melihat File
@@ -1,125 +1,127 @@
.. meta::
:description: This chapter describes a set of best practices designed to help developers optimize the performance of HIP-capable GPU architectures.
:description: This chapter describes a set of best practices designed to help
developers optimize the performance of HIP-capable GPU architectures.
:keywords: AMD, ROCm, HIP, CUDA, performance, guidelines
*******************************************************************************
Performance Guidelines
Performance guidelines
*******************************************************************************
The AMD HIP Performance Guidelines are a set of best practices designed to help
developers optimize the performance of AMD GPUs. They cover established
parallelization and optimization techniques, coding metaphors, and idioms that
can greatly simplify programming for HIP-capable GPU architectures.
The AMD HIP performance guidelines are a set of best practices designed to help
you optimize the application performance on AMDGPUs. The guidelines discuss
established parallelization and optimization techniques to improve the
application performance on HIP-capable GPU architectures.
By following four main cornerstones, we can exploit the performance
optimization potential of HIP.
Here are the four main cornerstones to help you exploit HIP's performance
optimization potential:
- parallel execution
- memory usage optimization
- optimization for maximum throughput
- minimizing memory thrashing
- Parallel execution
- Memory bandwidth usage optimization
- Maximum throughput optimization
- Memory thrashing minimization
In the following chapters, we will show you their benefits and how to use them
effectively.
This document discusses the usage and benefits of these cornerstones in detail.
.. _parallel execution:
Parallel execution
==================
================================================================================
For optimal use, the application should reveal and efficiently imply as much
parallelism as possible to keep all system components active.
For optimal use and to keep all system components busy, the application must
reveal and efficiently provide as much parallelism as possible. The parallelism
can be performed at the application level, device level, and multiprocessor
level.
Application level
-----------------
--------------------------------------------------------------------------------
The application should optimize parallel execution across the host and devices
using asynchronous calls and streams. Workloads should be assigned based on
efficiency: serial to the host, parallel to the devices.
To enable parallel execution of the application across the host and devices, use
asynchronous calls and streams. Assign workloads based on efficiency: serial to
the host or parallel to the devices.
For parallel workloads, when threads need to synchronize to share data, if they
belong to the same block, they should use ``__syncthreads()`` (see:
:ref:`synchronization functions`) within the same kernel invocation. If they
belong to different blocks, they must use global memory with two separate
kernel invocations. The latter should be minimized as it adds overhead.
For parallel workloads, when threads belonging to the same block need to
synchronize to share data, use :cpp:func:`__syncthreads()` (see:
:ref:`synchronization functions`) within the same kernel invocation. For threads
belonging to different blocks, use global memory with two separate
kernel invocations. It is recommended to avoid the latter approach as it adds
overhead.
Device level
------------
--------------------------------------------------------------------------------
Device-level optimization primarily involves maximizing parallel execution
across the multiprocessors of the device. This can be achieved by executing
multiple kernels concurrently on a device. The management of these kernels is
facilitated by streams, which allow for the overlapping of computation and data
transfers, enhancing performance. The aim is to keep all multiprocessors busy
by executing enough kernels concurrently. However, launching too many kernels
can lead to resource contention, so a balance must be found for optimal
performance. This approach helps in achieving maximum utilization of the
resources of the device.
Device level optimization primarily involves maximizing parallel execution
across the multiprocessors on the device. You can achieve device level
optimization by executing multiple kernels concurrently on a device. To enhance
performance, the management of these kernels is facilitated by streams, which
allows overlapping of computation and data transfers. This approach aims at
keeping all multiprocessors busy by executing enough kernels concurrently.
However, launching too many kernels can lead to resource contention, hence a
balance must be found for optimal performance. The device level optimization
helps in achieving maximum utilization of the device resources.
Multiprocessor level
--------------------
--------------------------------------------------------------------------------
Multiprocessor-level optimization involves maximizing parallel execution within
each multiprocessor on a device. Each multiprocessor can execute a number of
threads concurrently, and the total number of threads that can run in parallel
is determined by the number of concurrent threads each multiprocessor can
handle.
Multiprocessor level optimization involves maximizing parallel execution within
each multiprocessor on a device. The key to multiprocessor level optimization
is to efficiently utilize the various functional units within a multiprocessor.
For example, ensuring a sufficient number of resident warps, so that every clock
cycle has an instruction from a warp is ready for execution. This instruction
could either be another independent instruction of the same warp, which exploits
:ref:`instruction level optimization <instruction optimization>`, or more
commonly an instruction of another warp, which exploits thread-level parallelism.
The key to multiprocessor-level optimization is to efficiently utilize the
various functional units within a multiprocessor. This can be achieved by
ensuring a sufficient number of resident warps, as at every instruction issue
time, a warp scheduler selects an instruction that is ready to execute. This
instruction can be another independent instruction of the same warp, exploiting
:ref:`instruction optimization`, or more commonly an instruction of another warp,
exploiting thread-level parallelism.
In comparison, device-level optimization focuses on the device as a whole,
aiming to keep all multiprocessors busy by executing enough kernels
concurrently. Both levels of optimization are crucial for achieving maximum
performance. They work together to ensure efficient utilization of the
resources of the GPU, from the individual multiprocessors to the device as a
whole.
On the other hand, device level optimization focuses on the device as a whole,
aiming at keeping all multiprocessors busy by executing enough kernels
concurrently. Both multiprocessor and device levels of optimization are crucial
for achieving maximum performance. They work together to ensure efficient
utilization of the GPU resources, ranging from individual multiprocessors to the
device as a whole.
.. _memory optimization:
Memory optimization
===================
Memory throughput optimization
================================================================================
The first step in maximizing memory throughput is to minimize low-bandwidth
data transfers. This involves reducing data transfers between the host and the
device, as these have lower bandwidth than transfers between global memory and
the device.
data transfers between the host and the device.
Additionally, data transfers between global memory and the device should be
minimized by maximizing the use of on-chip memory: shared memory and caches.
Shared memory acts as a user-managed cache, where the application explicitly
allocates and accesses it. A common programming pattern is to stage data from
device memory into shared memory. This involves each thread of a block loading
data from device memory to shared memory, synchronizing with all other threads
of the block, processing the data in shared memory, synchronizing again if
necessary, and writing the results back to device global memory.
Additionally, maximize the use of on-chip memory, that is, shared memory and
caches, and minimize transfers with global memory. Shared memory acts as a
user-managed cache explicitly allocated and accessed by the application. A
common programming pattern is to stage data from device memory into shared
memory. The staging of data from the device to shared memory involves the
following steps:
1. Each thread of a block loading data from device memory to shared memory.
2. Synchronizing with all other threads of the block.
3. Processing the data stored in shared memory.
4. Synchronizing again if necessary.
5. Writing the results back to the device global memory.
For some applications, a traditional hardware-managed cache is more appropriate
to exploit data locality. On devices of certain compute capabilities, the same
on-chip memory is used for both L1 and shared memory, and the amount dedicated
to each is configurable for each kernel call.
for exploiting data locality.
Finally, the throughput of memory accesses by a kernel can vary significantly
depending on the access pattern for each type of memory. Therefore, the next
step in maximizing memory throughput is to organize memory accesses as
optimally as possible. This is especially important for global memory accesses,
as global memory bandwidth is low compared to available on-chip bandwidths and
arithmetic instruction throughput. Thus, non-optimal global memory accesses
generally have a high impact on performance.
In conclusion, the throughput of memory accesses by a kernel can vary
significantly depending on the access pattern. Therefore, the next step in
maximizing memory throughput is to organize memory accesses as optimally as
possible. This is especially important for global memory accesses, as global
memory bandwidth is low compared to available on-chip bandwidths and arithmetic
instruction throughput. Thus, non-optimal global memory accesses generally have
a high impact on performance.
The memory throughput optimization techniques are further discussed in detail in
the following sections.
Data Transfer
-------------
.. _data transfer:
Applications should aim to minimize data transfers between the host and the
device. This can be achieved by moving more computations from the host to the
device, even if it means running kernels that do not fully utilize the
parallelism for device. Intermediate data structures can be created, used,
and discarded in device memory without being mapped or copied to host memory.
Data transfer
--------------------------------------------------------------------------------
To minimize data transfers between the host and the device, applications should
move more computations from the host to the device, even at the cost of running
kernels that don't fully utilize parallelism for the device. Intermediate data
structures should be created, used, and discarded in device memory without being
mapped or copied to host memory.
Batching small transfers into a single large transfer can improve performance
due to the overhead associated with each transfer. On systems with a front-side
@@ -129,173 +131,185 @@ When using mapped page-locked memory, there is no need to allocate device
memory or explicitly copy data between device and host memory. Data transfers
occur implicitly each time the kernel accesses the mapped memory. For optimal
performance, these memory accesses should be coalesced, similar to global
memory accesses.
memory accesses. The process where threads in a warp access sequential memory
locations is known as coalesced memory access, which can enhance memory data
transfer efficiency.
On integrated systems where device and host memory are physically the same,
any copy operation between host and device memory is unnecessary, and mapped
page-locked memory should be used instead. Applications can check if a device
is integrated by querying the integrated device property.
On integrated systems where device and host memory are physically the same, no
copy operation between host and device memory is required and hence mapped
page-locked memory should be used instead. To check if the device is integrated,
applications can query the integrated device property.
.. _device memory access:
Device Memory Access
--------------------
Device memory access
---------------------
Memory access instructions may be repeated due to the spread of memory
Memory access instructions might be repeated due to the spread of memory
addresses across warp threads. The impact on throughput varies with memory type
and is generally reduced when addresses are more scattered, especially in
global memory.
Device memory is accessed via 32-, 64-, or 128-byte transactions that must be
naturally aligned. Maximizing memory throughput involves coalescing memory
accesses of threads within a warp into minimal transactions, following optimal
access patterns, using properly sized and aligned data types, and padding data
when necessary.
naturally aligned.
Maximizing memory throughput involves:
Global memory instructions support reading or writing data of specific sizes
(1, 2, 4, 8, or 16 bytes) that are naturally aligned. If the size and alignment
requirements are not met, it leads to multiple instructions, reducing
performance. Therefore, using data types that meet these requirements, ensuring
alignment for structures, and maintaining alignment for all values or arrays is
crucial for correct results and optimal performance.
- Coalescing memory accesses of threads within a warp into minimal transactions.
- Following optimal access patterns.
- Using properly sized and aligned data types.
- Padding data when necessary.
Global memory instructions support reading or writing data of specific sizes (1,
2, 4, 8, or 16 bytes) that are naturally aligned. Not meeting the size and
alignment requirements leads to multiple instructions, which reduces
performance. Therefore, for correct results and optimal performance:
- Use data types that meet these requirements
- Ensure alignment for structures
- Maintain alignment for all values or arrays.
Threads often access 2D arrays at an address calculated as
``BaseAddress + xIndex + width * yIndex``. For efficient memory access, the
array and thread block widths should be multiples of the warp size. If the
array width is not a multiple of the warp size, it is usually more efficient to
allocate it with a width rounded up to the nearest multiple and pad the rows
accordingly.
allocate the array with a width rounded up to the nearest multiple and pad the
rows accordingly.
Local memory is used for certain automatic variables, such as arrays with
non-constant indices, large structures or arrays, and any variable when the
non-constant indices, large structures of arrays, and any variable where the
kernel uses more registers than available. Local memory resides in device
memory, leading to high latency and low bandwidth similar to global memory
accesses. However, it is organized for consecutive 32-bit words to be accessed
by consecutive thread IDs, allowing full coalescing when all threads in a warp
access the same relative address.
memory, which leads to high latency and low bandwidth, similar to global memory
accesses. However, the local memory is organized for consecutive 32-bit words to
be accessed by consecutive thread IDs, which allows full coalescing when all
threads in a warp access the same relative address.
Shared memory, located on-chip, provides higher bandwidth and lower latency
than local or global memory. It is divided into banks that can be
simultaneously accessed, boosting bandwidth. However, bank conflicts, where two
addresses fall in the same bank, lead to serialized access and decreased
throughput. Therefore, understanding how memory addresses map to banks and
scheduling requests to minimize conflicts is crucial for optimal performance.
Shared memory is located on-chip and provides higher bandwidth and lower latency
than local or global memory. It is divided into banks that can be simultaneously
accessed, which boosts bandwidth. However, bank conflicts, where two addresses
fall in the same bank, lead to serialized access and decreased throughput.
Therefore, understanding how memory addresses map to banks and scheduling
requests to minimize conflicts is crucial for optimal performance.
Constant memory is in device memory and cached in the constant cache. Requests
are split based on different memory addresses, affecting throughput, and are
serviced at the throughput of the constant cache for cache hits, or the
throughput of the device memory otherwise.
Constant memory is in the device memory and cached in the constant cache.
Requests are split based on different memory addresses and are serviced based
either on the throughput of the constant cache for cache hits or on the
throughput of the device memory otherwise. This splitting of requests affects
throughput.
Texture and surface memory are stored in device memory and cached in texture
cache. This setup optimizes 2D spatial locality, leading to better performance
for threads reading close 2D addresses. Reading device memory through texture
or surface fetching can be advantageous, offering higher bandwidth for local
texture fetches or surface reads, offloading addressing calculations,
allowing data broadcasting, and optional conversion of 8-bit and 16-bit integer
input data to 32-bit floating-point values on-the-fly.
Texture and surface memory are stored in the device memory and cached in the
texture cache. This setup optimizes 2D spatial locality, which leads to better
performance for threads reading close 2D addresses.
Reading device memory through texture or surface fetching provides the following
advantages:
- Higher bandwidth for local texture fetches or surface reads.
- Offloading addressing calculation.
- Data broadcasting.
- Optional conversion of 8-bit and 16-bit integer input data to 32-bit
floating-point values on the fly.
.. _instruction optimization:
Optimization for maximum instruction throughput
===============================================
================================================================================
To maximize instruction throughput:
- minimize low throughput arithmetic instructions
- minimize divergent warps inflicted by control flow instructions
- minimize the number of instruction as possible
- maximize instruction parallelism
- Minimize low throughput arithmetic instructions.
- Minimize divergent warps inflicted by flow control instructions.
- Maximize instruction parallelism.
These techniques are discussed in detail in the following sections.
Arithmetic instructions
-----------------------
--------------------------------------------------------------------------------
The type and complexity of arithmetic operations can significantly impact the
performance of your application. We are highlighting some hints how to maximize
it.
Using efficient operations: Some arithmetic operations are more costly than
others. For example, multiplication is typically faster than division, and
integer operations are usually faster than floating-point operations,
especially with double-precision.
Use efficient operations: Some arithmetic operations are costlier than others.
For example, multiplication is typically faster than division, and integer
operations are usually faster than floating-point operations, especially with
double precision.
Minimizing low-throughput instructions: This might involve trading precision
for speed when it does not affect the final result. For instance, consider
using single-precision arithmetic instead of double-precision.
Minimize low-throughput instructions: This might involve trading precision for
speed when it does not affect the final result. For instance, consider using
single-precision arithmetic instead of double-precision.
Leverage intrinsic functions: Intrinsic functions are pre-defined functions
Leverage intrinsic functions: Intrinsic functions are predefined functions
available in HIP that can often be executed faster than equivalent arithmetic
operations (subject to some input or accuracy restrictions). They can help
optimize performance by replacing more complex arithmetic operations.
Avoiding divergent warps: Divergent warps occur when threads within the same
warp follow different execution paths. This can happen due to conditional
statements that lead to different arithmetic operations being performed by
different threads. Divergent warps can significantly reduce instruction
throughput, so try to structure your code to minimize divergence.
Optimize memory access: The memory access efficiency can impact the speed of
arithmetic operations. See: :ref:`device memory access`.
Optimizing memory access: The efficiency of memory access can impact the speed
of arithmetic operations. Coalesced memory access, where threads in a warp
access consecutive memory locations, can improve memory throughput and thus
the speed of arithmetic operations.
Maximizing instruction parallelism: Some GPU architectures could issue parallel
independent instructions simultaneously, for example integer and floating
point, or two operations with independent inputs and outputs. Mostly this is a
work for compiler, but expressing parallelism in the code explicitly can
improve instructions throughput.
.. _control flow instructions:
Control flow instructions
-------------------------
--------------------------------------------------------------------------------
Flow control instructions (``if``, ``else``, ``for``, ``do``, ``while``,
Control flow instructions (``if``, ``else``, ``for``, ``do``, ``while``,
``break``, ``continue``, ``switch``) can impact instruction throughput by
causing threads within a warp to diverge and follow different execution paths.
To optimize performance, control conditions should be written to minimize
divergent warps. For example, when the control condition depends on
(``threadIdx`` / ``warpSize``), no warp diverges. The compiler may optimize
loops or short if or switch blocks using branch predication, preventing warp
divergence. With branch predication, instructions associated with a false
predicate are scheduled but not executed, avoiding unnecessary operations.
To optimize performance, write control conditions to minimize divergent warps.
For example, when the control condition depends on ``threadIdx`` or ``warpSize``,
warp doesn't diverge. The compiler might optimize loops, short ifs, or switch
blocks using branch predication, which prevents warp divergence. With branch
predication, instructions associated with a false predicate are scheduled but
not executed, which avoids unnecessary operations.
Avoiding divergent warps
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Warps diverge when threads within the same warp follow different execution paths.
This is caused by conditional statements that lead to different arithmetic
operations being performed by different threads. Divergent warps can
significantly reduce instruction throughput, so it is advisable to structure
your code to minimize divergence.
Synchronization
---------------
--------------------------------------------------------------------------------
Synchronization ensures that all threads within a block have completed their
Synchronization ensures that all threads within a block complete their
computations and memory accesses before moving forward, which is critical when
threads are dependent on the results of other threads. However,
synchronization can also lead to performance overhead, as it requires threads
to wait, potentially leading to idle GPU resources.
threads depend on other thread results. However, synchronization can also cause
performance overhead, as it needs the threads to wait, which might lead to idle
GPU resources.
``__syncthreads()`` is used to synchronize all threads in a block, ensuring
that all threads have reached the same point in the code and that shared memory
is visible to all threads after the point of synchronization.
To synchronize all threads in a block, use :cpp:func:`__syncthreads()`.
:cpp:func:`__syncthreads()` ensures that, all threads reach the same point in
the code and can access shared memory after reaching that point.
An alternative way to synchronize is using streams. Different streams can
execute commands out of order with respect to one another or concurrently. This
allows for more fine-grained control over the execution order of commands,
which can be beneficial in certain scenarios.
An alternative way to synchronize is to use streams. Different streams can
execute commands either without following a specific order or concurrently. This
is why streams allow more fine-grained control over the execution order of
commands, which can be beneficial in certain scenarios.
Minimizing memory thrashing
===========================
================================================================================
Applications frequently allocating and freeing memory may experience slower
allocation calls over time. This is expected as memory is released back to the
operating system. To optimize performance in such scenarios, consider some
recommendations:
Applications frequently allocating and freeing memory might experience slower
allocation calls over time as memory is released back to the operating system.
To optimize performance in such scenarios, follow these guidelines:
- avoid allocating all available memory with ``hipMalloc`` / ``hipHostMalloc``,
as this immediately reserves memory and can block other applications from
using it. This could strain the operating system schedulers or even prevent
other applications from running on the same GPU.
- aim to allocate memory in suitably sized blocks early in the lifecycle of the
application and deallocate only when the application no longer needs it.
Minimize the number of ``hipMalloc`` and ``hipFree`` calls in your
application, particularly in areas critical to performance.
- if an application is unable to allocate sufficient device memory, consider
resorting to other memory types such as ``hipHostMalloc`` or
``hipMallocManaged``. While these may not offer the same performance, they
can allow the application to continue running.
- For supported platforms, ``hipMallocManaged`` allows for oversubscription.
With the right memory advise policies, it can maintain most, if not all, of
the performance of ``hipMalloc``. ``hipMallocManaged`` does not require an
allocation to be resident until it is needed or prefetched, easing the load
on the operating system schedulers and facilitating multi-tenant scenarios.
- Avoid allocating all available memory with :cpp:func:`hipMalloc` or
:cpp:func:`hipHostMalloc`, as this immediately reserves memory and might
prevent other applications from using it. This behavior could strain the
operating system schedulers or prevent other applications from running on the
same GPU.
- Try to allocate memory in suitably sized blocks early in the application's
lifecycle and deallocate only when the application no longer needs it.
Minimize the number of :cpp:func:`hipMalloc` and :cpp:func:`hipFree` calls in
your application, particularly in performance-critical areas.
- Consider resorting to other memory types such as :cpp:func:`hipHostMalloc` or
:cpp:func:`hipMallocManaged`, if an application can't allocate sufficient
device memory. While the other memory types might not offer similar
performance, they allow the application to continue running.
- For supported platforms, use :cpp:func:`hipMallocManaged`, as it allows
oversubscription. With the right policies, :cpp:func:`hipMallocManaged` can
maintain most, if not all, :cpp:func:`hipMalloc` performance.
:cpp:func:`hipMallocManaged` doesn't require an allocation to be resident
until it is needed or prefetched, which eases the load on the operating
system's schedulers and facilitates multitenant scenarios.
projects/hip/docs/how-to/programming_manual.md
+3 -3
Melihat File
@@ -1,4 +1,4 @@
# HIP Programming Manual
# HIP programming manual
## Host Memory
@@ -140,13 +140,13 @@ HIP now supports runtime compilation (HIP RTC), the usage of which will provide
HIP RTC APIs accept HIP source files in character string format as input parameters and create handles of programs by compiling the HIP source files without spawning separate processes.
For more details on HIP RTC APIs, refer to [HIP Runtime API Reference](https://rocm.docs.amd.com/projects/HIP/en/latest/doxygen/html/index.html).
For more details on HIP RTC APIs, refer to [HIP Runtime API Reference](../doxygen/html/index).
For Linux developers, the link [here](https://github.com/ROCm/hip-tests/blob/develop/samples/2_Cookbook/23_cmake_hiprtc/saxpy.cpp) shows an example how to program HIP application using runtime compilation mechanism, and a detailed [HIP RTC programming guide](./hip_rtc) is also available.
## HIP Graph
HIP graph is supported. For more details, refer to the HIP API Guide.
HIP graphs are supported. For more details, refer to the [HIP API Guide](../doxygen/html/group___graph) or the [how-to section for HIP graphs](../how-to/hipgraph).
## Device-Side Malloc
projects/hip/docs/how-to/stream_ordered_allocator.rst
+578
Melihat File
@@ -0,0 +1,578 @@
.. meta::
:description:
:keywords: stream, memory allocation, SOMA, stream ordered memory allocator
*******************************************************************************
Stream Ordered Memory Allocator
*******************************************************************************
The Stream Ordered Memory Allocator (SOMA) is part of the HIP runtime API. SOMA provides an asynchronous memory allocation mechanism with stream-ordering semantics. You can use SOMA to allocate and free memory in stream order, which ensures that all asynchronous accesses occur between the stream executions of allocation and deallocation. Compliance with stream order prevents use-before-allocation or use-after-free errors, which helps to avoid an undefined behavior.
Advantages of SOMA:
- Efficient reuse: Enables efficient memory reuse across streams, which reduces unnecessary allocation overhead.
- Fine-grained control: Allows you to set attributes and control caching behavior for memory pools.
- Inter-process sharing: Enables secure sharing of allocations between processes.
- Optimizations: Allows driver to optimize based on its awareness of SOMA and other stream management APIs.
Disadvantages of SOMA:
- Temporal constraints: Requires you to adhere strictly to stream order to avoid errors.
- Complexity: Involves memory management in stream order, which can be intricate.
- Learning curve: Requires you to put additional efforts to understand and utilize SOMA effectively.
Using SOMA
=====================================
You can allocate memory using ``hipMallocAsync()`` with stream-ordered
semantics. This restricts the asynchronous access to the memory between the stream executions of the allocation and deallocation. Accessing
memory if the compliant memory accesses won't overlap
temporally. ``hipFreeAsync()`` frees memory from the pool with stream-ordered
semantics.
Here is how to use stream ordered memory allocation:
.. tab-set::
.. tab-item:: Stream Ordered Memory Allocation
.. code-block:: cpp
#include <iostream>
#include <hip/hip_runtime.h>
// Kernel to perform some computation on allocated memory.
__global__ void myKernel(int* data, size_t numElements) {
int tid = threadIdx.x + blockIdx.x * blockDim.x;
if (tid < numElements) {
data[tid] = tid * 2;
}
}
int main() {
// Initialize HIP.
hipInit(0);
// Stream 0.
constexpr hipStream_t streamId = 0;
// Allocate memory with stream ordered semantics.
constexpr size_t numElements = 1024;
int* devData;
hipMallocAsync(&devData, numElements * sizeof(*devData), streamId);
// Launch the kernel to perform computation.
dim3 blockSize(256);
dim3 gridSize((numElements + blockSize.x - 1) / blockSize.x);
myKernel<<<gridSize, blockSize>>>(devData, numElements);
// Copy data back to host.
int* hostData = new int[numElements];
hipMemcpy(hostData, devData, numElements * sizeof(*devData), hipMemcpyDeviceToHost);
// Print the array.
for (size_t i = 0; i < numElements; ++i) {
std::cout << "Element " << i << ": " << hostData[i] << std::endl;
}
// Free memory with stream ordered semantics.
hipFreeAsync(devData, streamId);
delete[] hostData;
// Synchronize to ensure completion.
hipDeviceSynchronize();
return 0;
}
.. tab-item:: Ordinary Allocation
.. code-block:: cpp
#include <iostream>
#include <hip/hip_runtime.h>
// Kernel to perform some computation on allocated memory.
__global__ void myKernel(int* data, size_t numElements) {
int tid = threadIdx.x + blockIdx.x * blockDim.x;
if (tid < numElements) {
data[tid] = tid * 2;
}
}
int main() {
// Initialize HIP.
hipInit(0);
// Allocate memory.
constexpr size_t numElements = 1024;
int* devData;
hipMalloc(&devData, numElements * sizeof(*devData));
// Launch the kernel to perform computation.
dim3 blockSize(256);
dim3 gridSize((numElements + blockSize.x - 1) / blockSize.x);
myKernel<<<gridSize, blockSize>>>(devData, numElements);
// Copy data back to host.
int* hostData = new int[numElements];
hipMemcpy(hostData, devData, numElements * sizeof(*devData), hipMemcpyDeviceToHost);
// Print the array.
for (size_t i = 0; i < numElements; ++i) {
std::cout << "Element " << i << ": " << hostData[i] << std::endl;
}
// Free memory.
hipFree(devData);
delete[] hostData;
// Synchronize to ensure completion.
hipDeviceSynchronize();
return 0;
}
For more details, see :ref:`stream_ordered_memory_allocator_reference`.
Memory pools
============
Memory pools provide a way to manage memory with stream-ordered behavior while ensuring proper synchronization and avoiding memory access errors. Division of a single memory system into separate pools facilitates querying the access path properties for each partition. Memory pools are used for host memory, device memory, and unified memory.
Set pools
---------
The ``hipMallocAsync()`` function uses the current memory pool and also provides the opportunity to create and access different pools using ``hipMemPoolCreate()`` and ``hipMallocFromPoolAsync()`` functions respectively.
Unlike NVIDIA CUDA, where stream-ordered memory allocation can be implicit, ROCm HIP is explicit. This requires managing memory allocation for each stream in HIP while ensuring precise control over memory usage and synchronization.
.. code-block:: cpp
#include <iostream>
#include <hip/hip_runtime.h>
// Kernel to perform some computation on allocated memory.
__global__ void myKernel(int* data, size_t numElements) {
int tid = threadIdx.x + blockIdx.x * blockDim.x;
if (tid < numElements) {
data[tid] = tid * 2;
}
}
int main() {
// Create a stream.
hipStream_t stream;
hipStreamCreate(&stream);
// Create a memory pool with default properties.
hipMemPoolProps poolProps = {};
poolProps.allocType = hipMemAllocationTypePinned;
poolProps.handleTypes = hipMemHandleTypePosixFileDescriptor;
poolProps.location.type = hipMemLocationTypeDevice;
poolProps.location.id = 0; // Assuming device 0.
hipMemPool_t memPool;
hipMemPoolCreate(&memPool, &poolProps);
// Allocate memory from the pool asynchronously.
constexpr size_t numElements = 1024;
int* devData = nullptr;
hipMallocFromPoolAsync(&devData, numElements * sizeof(*devData), memPool, stream);
// Define grid and block sizes.
dim3 blockSize(256);
dim3 gridSize((numElements + blockSize.x - 1) / blockSize.x);
// Launch the kernel to perform computation.
myKernel<<<gridSize, blockSize, 0, stream>>>(devData, numElements);
// Synchronize the stream.
hipStreamSynchronize(stream);
// Copy data back to host.
int* hostData = new int[numElements];
hipMemcpy(hostData, devData, numElements * sizeof(*devData), hipMemcpyDeviceToHost);
// Print the array.
for (size_t i = 0; i < numElements; ++i) {
std::cout << "Element " << i << ": " << hostData[i] << std::endl;
}
// Free the allocated memory.
hipFreeAsync(devData, stream);
// Synchronize the stream again to ensure all operations are complete.
hipStreamSynchronize(stream);
// Destroy the memory pool and stream.
hipMemPoolDestroy(memPool);
hipStreamDestroy(stream);
// Free host memory.
delete[] hostData;
return 0;
}
Trim pools
----------
The memory allocator allows you to allocate and free memory in stream order. To control memory usage, set the release threshold attribute using ``hipMemPoolAttrReleaseThreshold``. This threshold specifies the amount of reserved memory in bytes to hold onto.
.. code-block:: cpp
uint64_t threshold = UINT64_MAX;
hipMemPoolSetAttribute(memPool, hipMemPoolAttrReleaseThreshold, &threshold);
When the amount of memory held in the memory pool exceeds the threshold, the allocator tries to release memory back to the operating system during the next call to stream, event, or context synchronization.
To improve performance, it is a good practice to adjust the memory pool size using ``hipMemPoolTrimTo()``. It helps to reclaim memory from an excessive memory pool, which optimizes memory usage for your application.
.. code-block:: cpp
#include <hip/hip_runtime.h>
#include <iostream>
int main() {
hipMemPool_t memPool;
hipDevice_t device = 0; // Specify the device index.
// Initialize the device.
hipSetDevice(device);
// Get the default memory pool for the device.
hipDeviceGetDefaultMemPool(&memPool, device);
// Allocate memory from the pool (e.g., 1 MB).
size_t allocSize = 1 * 1024 * 1024;
void* ptr;
hipMalloc(&ptr, allocSize);
// Free the allocated memory.
hipFree(ptr);
// Trim the memory pool to a specific size (e.g., 512 KB).
size_t newSize = 512 * 1024;
hipMemPoolTrimTo(memPool, newSize);
// Clean up.
hipMemPoolDestroy(memPool);
std::cout << "Memory pool trimmed to " << newSize << " bytes." << std::endl;
return 0;
}
Resource usage statistics
-------------------------
Resource usage statistics help in optimization. Here is the list of pool attributes used to query memory usage:
- ``hipMemPoolAttrReservedMemCurrent``: Returns the total physical GPU memory currently held in the pool.
- ``hipMemPoolAttrUsedMemCurrent``: Returns the total size of all the memory allocated from the pool.
- ``hipMemPoolAttrReservedMemHigh``: Returns the total physical GPU memory held in the pool since the last reset.
- ``hipMemPoolAttrUsedMemHigh``: Returns the total size of all the memory allocated from the pool since the last reset.
To reset these attributes to the current value, use ``hipMemPoolSetAttribute()``.
.. code-block:: cpp
#include <iostream>
#include <hip/hip_runtime.h>
// Sample helper functions for getting the usage statistics in bulk.
struct usageStatistics {
uint64_t reservedMemCurrent;
uint64_t reservedMemHigh;
uint64_t usedMemCurrent;
uint64_t usedMemHigh;
};
void getUsageStatistics(hipMemPool_t memPool, struct usageStatistics *statistics) {
hipMemPoolGetAttribute(memPool, hipMemPoolAttrReservedMemCurrent, &statistics->reservedMemCurrent);
hipMemPoolGetAttribute(memPool, hipMemPoolAttrReservedMemHigh, &statistics->reservedMemHigh);
hipMemPoolGetAttribute(memPool, hipMemPoolAttrUsedMemCurrent, &statistics->usedMemCurrent);
hipMemPoolGetAttribute(memPool, hipMemPoolAttrUsedMemHigh, &statistics->usedMemHigh);
}
// Resetting the watermarks resets them to the current value.
void resetStatistics(hipMemPool_t memPool) {
uint64_t value = 0;
hipMemPoolSetAttribute(memPool, hipMemPoolAttrReservedMemHigh, &value);
hipMemPoolSetAttribute(memPool, hipMemPoolAttrUsedMemHigh, &value);
}
int main() {
hipMemPool_t memPool;
hipDevice_t device = 0; // Specify the device index.
// Initialize the device.
hipSetDevice(device);
// Get the default memory pool for the device.
hipDeviceGetDefaultMemPool(&memPool, device);
// Allocate memory from the pool (e.g., 1 MB).
size_t allocSize = 1 * 1024 * 1024;
void* ptr;
hipMalloc(&ptr, allocSize);
// Free the allocated memory.
hipFree(ptr);
// Trim the memory pool to a specific size (e.g., 512 KB).
size_t newSize = 512 * 1024;
hipMemPoolTrimTo(memPool, newSize);
// Get and print usage statistics before resetting.
usageStatistics statsBefore;
getUsageStatistics(memPool, &statsBefore);
std::cout << "Before resetting statistics:" << std::endl;
std::cout << "Reserved Memory Current: " << statsBefore.reservedMemCurrent << " bytes" << std::endl;
std::cout << "Reserved Memory High: " << statsBefore.reservedMemHigh << " bytes" << std::endl;
std::cout << "Used Memory Current: " << statsBefore.usedMemCurrent << " bytes" << std::endl;
std::cout << "Used Memory High: " << statsBefore.usedMemHigh << " bytes" << std::endl;
// Reset the statistics.
resetStatistics(memPool);
// Get and print usage statistics after resetting.
usageStatistics statsAfter;
getUsageStatistics(memPool, &statsAfter);
std::cout << "After resetting statistics:" << std::endl;
std::cout << "Reserved Memory Current: " << statsAfter.reservedMemCurrent << " bytes" << std::endl;
std::cout << "Reserved Memory High: " << statsAfter.reservedMemHigh << " bytes" << std::endl;
std::cout << "Used Memory Current: " << statsAfter.usedMemCurrent << " bytes" << std::endl;
std::cout << "Used Memory High: " << statsAfter.usedMemHigh << " bytes" << std::endl;
// Clean up.
hipMemPoolDestroy(memPool);
return 0;
}
Memory reuse policies
---------------------
The allocator might reallocate memory as long as the compliant memory accesses will not to overlap temporally. To optimize the memory usage, disable or enable the following memory pool reuse policy attribute flags:
- ``hipMemPoolReuseFollowEventDependencies``: Checks event dependencies before allocating additional GPU memory.
- ``hipMemPoolReuseAllowOpportunistic``: Checks freed allocations to determine if the stream order semantic indicated by the free operation has been met.
- ``hipMemPoolReuseAllowInternalDependencies``: Manages reuse based on internal dependencies in runtime. If the driver fails to allocate and map additional physical memory, it searches for memory waiting for another stream's progress and reuses it.
Device accessibility for multi-GPU support
------------------------------------------
Allocations are initially accessible from the device where they reside.
Interprocess memory handling
=============================
Interprocess capable (IPC) memory pools facilitate efficient and secure sharing of GPU memory between processes.
To achieve interprocess memory sharing, you can use either :ref:`device pointer <device-pointer>` or :ref:`shareable handle <shareable-handle>`. Both provide allocator (export) and consumer (import) interfaces.
.. _device-pointer:
Device pointer
--------------
To export data to share a memory pool pointer directly between processes, use ``hipMemPoolExportPointer()``. It allows you to share a memory allocation with another process.
.. code-block:: cpp
#include <iostream>
#include <fstream>
#include <hip/hip_runtime.h>
#include <sys/stat.h>
int main() {
// Allocate memory.
void* devPtr;
hipMalloc(&devPtr, sizeof(int));
// Export the memory pool pointer.
hipMemPoolPtrExportData exportData;
hipError_t result = hipMemPoolExportPointer(&exportData, devPtr);
if (result != hipSuccess) {
std::cerr << "Error exporting memory pool pointer: " << hipGetErrorString(result) << std::endl;
return 1;
}
// Create a named pipe (FIFO).
const char* fifoPath = "/tmp/myfifo"; // Change this to a unique path.
mkfifo(fifoPath, 0666);
// Write the exported data to the named pipe.
std::ofstream fifoStream(fifoPath, std::ios::out | std::ios::binary);
fifoStream.write(reinterpret_cast<char*>(&exportData), sizeof(hipMemPoolPtrExportData));
fifoStream.close();
// Clean up.
hipFree(devPtr);
return 0;
}
To import a memory pool pointer directly from another process, use ``hipMemPoolImportPointer()``.
Here is how to read the pool exported in the preceding example:
.. code-block:: cpp
#include <iostream>
#include <fstream>
#include <hip/hip_runtime.h>
int main() {
// Considering that you have exported the memory pool pointer already.
// Now, let's simulate reading the exported data from a named pipe (FIFO).
const char* fifoPath = "/tmp/myfifo"; // Change this to a unique path.
std::ifstream fifoStream(fifoPath, std::ios::in | std::ios::binary);
if (!fifoStream.is_open()) {
std::cerr << "Error opening FIFO file: " << fifoPath << std::endl;
return 1;
}
// Read the exported data.
hipMemPoolPtrExportData importData;
fifoStream.read(reinterpret_cast<char*>(&importData), sizeof(hipMemPoolPtrExportData));
fifoStream.close();
if (fifoStream.fail()) {
std::cerr << "Error reading from FIFO file." << std::endl;
return 1;
}
// Create a memory pool with default properties.
hipMemPoolProps poolProps = {};
poolProps.allocType = hipMemAllocationTypePinned;
poolProps.handleTypes = hipMemHandleTypePosixFileDescriptor;
poolProps.location.type = hipMemLocationTypeDevice;
poolProps.location.id = 0; // Assuming device 0.
hipMemPool_t memPool;
hipMemPoolCreate(&memPool, &poolProps);
// Import the memory pool pointer.
void* importedDevPtr;
hipError_t result = hipMemPoolImportPointer(&importedDevPtr, memPool, &importData);
if (result != hipSuccess) {
std::cerr << "Error imported memory pool pointer: " << hipGetErrorString(result) << std::endl;
return 1;
}
// Now you can use the importedDevPtr for your computations.
// Clean up (free the memory).
hipFree(importedDevPtr);
return 0;
}
.. _shareable-handle:
Shareable handle
----------------
To export a memory pool pointer to a shareable handle, use ``hipMemPoolExportToSharedHandle()``. This handle could be a file descriptor or a handle obtained from another process. The exported handle contains information about the memory pool, such as size, location, and other relevant details.
.. code-block:: cpp
#include <iostream>
#include <fstream>
#include <hip/hip_runtime.h>
#include <sys/stat.h>
int main() {
// Create a memory pool with default properties.
hipMemPoolProps poolProps = {};
poolProps.allocType = hipMemAllocationTypePinned;
poolProps.handleTypes = hipMemHandleTypePosixFileDescriptor;
poolProps.location.type = hipMemLocationTypeDevice;
poolProps.location.id = 0; // Assuming device 0.
hipMemPool_t memPool;
hipError_t poolResult = hipMemPoolCreate(&memPool, &poolProps);
if (poolResult != hipSuccess) {
std::cerr << "Error creating memory pool: " << hipGetErrorString(poolResult) << std::endl;
return 1;
}
// Allocate memory from the memory pool.
void* devPtr;
hipMallocFromPoolAsync(&devPtr, sizeof(int), memPool, 0);
// Export the memory pool pointer.
int descriptor;
hipError_t result = hipMemPoolExportToShareableHandle(&descriptor, memPool, hipMemHandleTypePosixFileDescriptor, 0);
if (result != hipSuccess) {
std::cerr << "Error exporting memory pool pointer: " << hipGetErrorString(result) << std::endl;
return 1;
}
// Create a named pipe (FIFO).
const char* fifoPath = "/tmp/myfifo"; // Change this to a unique path.
mkfifo(fifoPath, 0666);
// Write the exported data to the named pipe.
std::ofstream fifoStream(fifoPath, std::ios::out | std::ios::binary);
fifoStream.write(reinterpret_cast<char*>(&descriptor), sizeof(int));
fifoStream.close();
// Clean up.
hipFree(devPtr);
hipMemPoolDestroy(memPool);
return 0;
}
To import and restore a memory pool pointer from a shareable handle, which could be a file descriptor or a handle obtained from another process, use ``hipMemPoolImportFromShareableHandle()``. The exported shareable handle data contains information about the memory pool, including its size, location, and other relevant details. Importing the handle provides a valid memory pointer to the same memory, which allows you to share memory across different contexts.
.. code-block:: cpp
#include <iostream>
#include <fstream>
#include <hip/hip_runtime.h>
int main() {
// Considering that you have exported the memory pool pointer already.
// Now, let's simulate reading the exported data from a named pipe (FIFO).
const char* fifoPath = "/tmp/myfifo"; // Change this to a unique path
std::ifstream fifoStream(fifoPath, std::ios::in | std::ios::binary);
if (!fifoStream.is_open()) {
std::cerr << "Error opening FIFO file: " << fifoPath << std::endl;
return 1;
}
// Read the exported data.
int descriptor;
fifoStream.read(reinterpret_cast<char*>(&descriptor), sizeof(int));
fifoStream.close();
if (fifoStream.fail()) {
std::cerr << "Error reading from FIFO file." << std::endl;
return 1;
}
// Import the memory pool.
hipMemPool_t memPool;
hipError_t result = hipMemPoolImportFromShareableHandle(&memPool, &descriptor, hipMemHandleTypePosixFileDescriptor, 0);
if (result != hipSuccess) {
std::cerr << "Error importing memory pool: " << hipGetErrorString(result) << std::endl;
return 1;
}
// Allocate memory from the imported memory pool.
void* importedDevPtr;
hipMallocFromPoolAsync(&importedDevPtr, sizeof(int), memPool, 0);
// Now you can use the importedDevPtr for your computations.
// Clean up (free the memory).
hipFree(importedDevPtr);
hipMemPoolDestroy(memPool);
return 0;
}
projects/hip/docs/how-to/unified_memory.rst
+577
Melihat File
@@ -0,0 +1,577 @@
.. meta::
:description: This chapter describes introduces Unified Memory (UM) and shows
how to use it in AMD HIP.
:keywords: AMD, ROCm, HIP, CUDA, unified memory, unified, memory, UM, APU
*******************************************************************************
Unified memory
*******************************************************************************
In conventional architectures, CPUs and GPUs have dedicated memory like Random
Access Memory (RAM) and Video Random Access Memory (VRAM). This architectural
design, while effective, can be limiting in terms of memory capacity and
bandwidth, as continuous memory copying is required to allow the processors to
access the appropriate data. New architectural features like Heterogeneous
System Architectures (HSA) and Unified Memory (UM) help avoid these limitations
and promise increased efficiency and innovation.
Unified memory
==============
Unified Memory is a single memory address space accessible from any processor
within a system. This setup simplifies memory management processes and enables
applications to allocate data that can be read or written by code running on
either CPUs or GPUs. The Unified memory model is shown in the following figure.
.. figure:: ../data/unified_memory/um.svg
AMD Accelerated Processing Unit (APU) is a typical example of a Unified Memory
Architecture. On a single die, a central processing unit (CPU) is combined
with an integrated graphics processing unit (iGPU), and both have access to a
high-bandwidth memory (HBM) module named Unified Memory. The CPU enables
high-performance, low-latency operations, while the GPU is optimized for high
throughput (data processed by unit time).
.. _unified memory system requirements:
System requirements
===================
Unified memory is supported on Linux by all modern AMD GPUs from the Vega
series onward. Unified memory management can be achieved with managed memory
allocation and, for the latest GPUs, with a system allocator.
The table below lists the supported allocators. The allocators are described in
the next section.
.. list-table:: Supported Unified Memory Allocators
:widths: 40, 25, 25, 25
:header-rows: 1
:align: center
* - Architecture
- ``hipMallocManaged()``
- ``__managed__``
- ``malloc()``
* - MI200, MI300 Series
- ✅
- ✅
- ✅ :sup:`1`
* - MI100
- ✅
- ✅
- ❌
* - RDNA (Navi) Series
- ✅
- ✅
- ❌
* - GCN5 (Vega) Series
- ✅
- ✅
- ❌
✅: **Supported**
❌: **Unsupported**
:sup:`1` Works only with ``XNACK=1``. First GPU access causes recoverable
page-fault. For more details, visit
`GPU memory <https://rocm.docs.amd.com/en/latest/conceptual/gpu-memory.html#xnack>`_.
.. _unified memory programming models:
Unified memory programming models
=================================
Showcasing various unified memory programming models, the model availability
depends on your architecture. For more information, see :ref:`unified memory
system requirements` and :ref:`checking unified memory management support`.
- **HIP managed memory allocation API**:
The ``hipMallocManaged()`` is a dynamic memory allocator available on
all GPUs with unified memory support. For more details, visit
:ref:`unified_memory_reference`.
- **HIP managed variables**:
The ``__managed__`` declaration specifier, which serves as its counterpart,
is supported on all modern AMD cards and can be utilized for static
allocation.
- **System allocation API**:
Starting with the AMD MI300 series, the ``malloc()`` system allocator allows
you to reserve unified memory. The system allocator is more versatile and
offers an easy transition from a CPU written C++ code to a HIP code as the
same system allocation API is used.
.. _checking unified memory management support:
Checking unified memory management support
------------------------------------------
Some device attributes can offer information about which :ref:`unified memory
programming models` are supported. The attribute value is 1 if the
functionality is supported, and 0 if it is not supported.
.. list-table:: Device attributes for unified memory management
:widths: 40, 60
:header-rows: 1
:align: center
* - attribute
- description
* - ``hipDeviceAttributeManagedMemory``
- unified addressing is supported
* - ``hipDeviceAttributeConcurrentManagedAccess``
- full managed memory support, concurrent access is supported
* - ``hipDeviceAttributePageableMemoryAccess``
- both managed and system memory allocation API is supported
The following examples show how to use device attributes:
.. code-block:: cpp
#include <hip/hip_runtime.h>
#include <iostream>
int main() {
int d;
hipGetDevice(&d);
int is_cma = 0;
hipDeviceGetAttribute(&is_cma, hipDeviceAttributeConcurrentManagedAccess, d);
std::cout << "HIP Managed Memory: "
<< (is_cma == 1 ? "is" : "NOT")
<< " supported" << std::endl;
return 0;
}
Example for unified memory management
-------------------------------------
The following example shows how to use unified memory management with
``hipMallocManaged()``, function, with ``__managed__`` attribute for static
allocation and standard ``malloc()`` allocation. For comparison, the Explicit
Memory Management example is presented in the last tab.
.. tab-set::
.. tab-item:: hipMallocManaged()
.. code-block:: cpp
:emphasize-lines: 12-15
#include <hip/hip_runtime.h>
#include <iostream>
// Addition of two values.
__global__ void add(int *a, int *b, int *c) {
*c = *a + *b;
}
int main() {
int *a, *b, *c;
// Allocate memory for a, b and c that is accessible to both device and host codes.
hipMallocManaged(&a, sizeof(*a));
hipMallocManaged(&b, sizeof(*b));
hipMallocManaged(&c, sizeof(*c));
// Setup input values.
*a = 1;
*b = 2;
// Launch add() kernel on GPU.
hipLaunchKernelGGL(add, dim3(1), dim3(1), 0, 0, a, b, c);
// Wait for GPU to finish before accessing on host.
hipDeviceSynchronize();
// Prints the result.
std::cout << *a << " + " << *b << " = " << *c << std::endl;
// Cleanup allocated memory.
hipFree(a);
hipFree(b);
hipFree(c);
return 0;
}
.. tab-item:: __managed__
.. code-block:: cpp
:emphasize-lines: 9-10
#include <hip/hip_runtime.h>
#include <iostream>
// Addition of two values.
__global__ void add(int *a, int *b, int *c) {
*c = *a + *b;
}
// Declare a, b and c as static variables.
__managed__ int a, b, c;
int main() {
// Setup input values.
a = 1;
b = 2;
// Launch add() kernel on GPU.
hipLaunchKernelGGL(add, dim3(1), dim3(1), 0, 0, &a, &b, &c);
// Wait for GPU to finish before accessing on host.
hipDeviceSynchronize();
// Prints the result.
std::cout << a << " + " << b << " = " << c << std::endl;
return 0;
}
.. tab-item:: malloc()
.. code-block:: cpp
:emphasize-lines: 12-15
#include <hip/hip_runtime.h>
#include <iostream>
// Addition of two values.
__global__ void add(int* a, int* b, int* c) {
*c = *a + *b;
}
int main() {
int* a, * b, * c;
// Allocate memory for a, b, and c.
a = (int*)malloc(sizeof(*a));
b = (int*)malloc(sizeof(*b));
c = (int*)malloc(sizeof(*c));
// Setup input values.
*a = 1;
*b = 2;
// Launch add() kernel on GPU.
hipLaunchKernelGGL(add, dim3(1), dim3(1), 0, 0, a, b, c);
// Wait for GPU to finish before accessing on host.
hipDeviceSynchronize();
// Prints the result.
std::cout << *a << " + " << *b << " = " << *c << std::endl;
// Cleanup allocated memory.
free(a);
free(b);
free(c);
return 0;
}
.. tab-item:: Explicit Memory Management
.. code-block:: cpp
:emphasize-lines: 17-24, 29-30
#include <hip/hip_runtime.h>
#include <iostream>
// Addition of two values.
__global__ void add(int *a, int *b, int *c) {
*c = *a + *b;
}
int main() {
int a, b, c;
int *d_a, *d_b, *d_c;
// Setup input values.
a = 1;
b = 2;
// Allocate device copies of a, b and c.
hipMalloc(&d_a, sizeof(*d_a));
hipMalloc(&d_b, sizeof(*d_b));
hipMalloc(&d_c, sizeof(*d_c));
// Copy input values to device.
hipMemcpy(d_a, &a, sizeof(*d_a), hipMemcpyHostToDevice);
hipMemcpy(d_b, &b, sizeof(*d_b), hipMemcpyHostToDevice);
// Launch add() kernel on GPU.
hipLaunchKernelGGL(add, dim3(1), dim3(1), 0, 0, d_a, d_b, d_c);
// Copy the result back to the host.
hipMemcpy(&c, d_c, sizeof(*d_c), hipMemcpyDeviceToHost);
// Cleanup allocated memory.
hipFree(d_a);
hipFree(d_b);
hipFree(d_c);
// Prints the result.
std::cout << a << " + " << b << " = " << c << std::endl;
return 0;
}
.. _using unified memory management:
Using unified memory management (UMM)
=====================================
Unified memory management (UMM) is a feature that can simplify the complexities
of memory management in GPU computing. It is particularly useful in
heterogeneous computing environments with heavy memory usage with both a CPU
and a GPU, which would require large memory transfers. Here are some areas
where UMM can be beneficial:
- **Simplification of Memory Management**:
UMM can help to simplify the complexities of memory management. This can make
it easier for developers to write code without worrying about memory
allocation and deallocation details.
- **Data Migration**:
UMM allows for efficient data migration between the host (CPU) and the device
(GPU). This can be particularly useful for applications that need to move
data back and forth between the device and host.
- **Improved Programming Productivity**:
As a positive side effect, UMM can reduce the lines of code, thereby
improving programming productivity.
In HIP, pinned memory allocations are coherent by default. Pinned memory is
host memory mapped into the address space of all GPUs, meaning that the pointer
can be used on both host and device. Using pinned memory instead of pageable
memory on the host can improve bandwidth.
While UMM can provide numerous benefits, it's important to be aware of the
potential performance overhead associated with UMM. You must thoroughly test
and profile your code to ensure it's the most suitable choice for your use
case.
.. _unified memory runtime hints:
Unified memory HIP runtime hints for the better performance
===========================================================
Unified memory HIP runtime hints can help improve the performance of your code if
you know your code's ability and infrastructure. Some hint techniques are
presented in this section.
The hint functions can set actions on a selected device, which can be
identified by ``hipGetDeviceProperties(&prop, device_id)``. There are two
special ``device_id`` values:
- ``hipCpuDeviceId`` = -1 means that the advised device is the CPU.
- ``hipInvalidDeviceId`` = -2 means that the device is invalid.
For the best performance, profile your application to optimize the
utilization of HIP runtime hints.
Data prefetching
----------------
Data prefetching is a technique used to improve the performance of your
application by moving data closer to the processing unit before it's actually
needed.
.. code-block:: cpp
:emphasize-lines: 20-23,31-32
// Addition of two values.
__global__ void add(int *a, int *b, int *c) {
*c = *a + *b;
}
int main() {
int *a, *b, *c;
int deviceId;
hipGetDevice(&deviceId); // Get the current device ID
// Allocate memory for a, b and c that is accessible to both device and host codes.
hipMallocManaged(&a, sizeof(*a));
hipMallocManaged(&b, sizeof(*b));
hipMallocManaged(&c, sizeof(*c));
// Setup input values.
*a = 1;
*b = 2;
// Prefetch the data to the GPU device.
hipMemPrefetchAsync(a, sizeof(*a), deviceId, 0);
hipMemPrefetchAsync(b, sizeof(*b), deviceId, 0);
hipMemPrefetchAsync(c, sizeof(*c), deviceId, 0);
// Launch add() kernel on GPU.
hipLaunchKernelGGL(add, dim3(1), dim3(1), 0, 0, a, b, c);
// Wait for GPU to finish before accessing on host.
hipDeviceSynchronize();
// Prefetch the result back to the CPU.
hipMemPrefetchAsync(c, sizeof(*c), hipCpuDeviceId, 0);
// Wait for the prefetch operations to complete.
hipDeviceSynchronize();
// Prints the result.
std::cout << *a << " + " << *b << " = " << *c << std::endl;
// Cleanup allocated memory.
hipFree(a);
hipFree(b);
hipFree(c);
return 0;
}
Remember to check the return status of ``hipMemPrefetchAsync()`` to ensure that
the prefetch operations are completed successfully.
Memory advice
-------------
The effectiveness of ``hipMemAdvise()`` comes from its ability to inform the
runtime system of the developer's intentions regarding memory usage. When the
runtime system has knowledge of the expected memory access patterns, it can
make better decisions about data placement and caching, leading to more
efficient execution of the application. However, the actual impact on
performance can vary based on the specific use case and the hardware
architecture.
For the description of ``hipMemAdvise()`` and the detailed list of advice,
visit the :ref:`unified_memory_reference`.
Here is the updated version of the example above with memory advice.
.. code-block:: cpp
:emphasize-lines: 17-26
#include <hip/hip_runtime.h>
#include <iostream>
// Addition of two values.
__global__ void add(int *a, int *b, int *c) {
*c = *a + *b;
}
int main() {
int *a, *b, *c;
// Allocate memory for a, b, and c accessible to both device and host codes.
hipMallocManaged(&a, sizeof(*a));
hipMallocManaged(&b, sizeof(*b));
hipMallocManaged(&c, sizeof(*c));
// Set memory advice for a, b, and c to be accessed by the CPU.
hipMemAdvise(a, sizeof(*a), hipMemAdviseSetPreferredLocation, hipCpuDeviceId);
hipMemAdvise(b, sizeof(*b), hipMemAdviseSetPreferredLocation, hipCpuDeviceId);
hipMemAdvise(c, sizeof(*c), hipMemAdviseSetPreferredLocation, hipCpuDeviceId);
// Additionally, set memory advice for a, b, and c to be read mostly from the device 0.
constexpr int device = 0;
hipMemAdvise(a, sizeof(*a), hipMemAdviseSetReadMostly, device);
hipMemAdvise(b, sizeof(*b), hipMemAdviseSetReadMostly, device);
hipMemAdvise(c, sizeof(*c), hipMemAdviseSetReadMostly, device);
// Setup input values.
*a = 1;
*b = 2;
// Launch add() kernel on GPU.
hipLaunchKernelGGL(add, dim3(1), dim3(1), 0, 0, a, b, c);
// Wait for GPU to finish before accessing on host.
hipDeviceSynchronize();
// Prints the result.
std::cout << *a << " + " << *b << " = " << *c << std::endl;
// Cleanup allocated memory.
hipFree(a);
hipFree(b);
hipFree(c);
return 0;
}
Memory range attributes
-----------------------
Memory Range attributes allow you to query attributes of a given memory range.
The ``hipMemRangeGetAttribute()`` is added to the example to query the
``hipMemRangeAttributeReadMostly`` attribute of the memory range pointed to by
``a``. The result is stored in ``attributeValue`` and then printed out.
For more details, visit the
:ref:`unified_memory_reference`.
.. code-block:: cpp
:emphasize-lines: 29-34
#include <hip/hip_runtime.h>
#include <iostream>
// Addition of two values.
__global__ void add(int *a, int *b, int *c) {
*c = *a + *b;
}
int main() {
int *a, *b, *c;
unsigned int attributeValue;
constexpr size_t attributeSize = sizeof(attributeValue);
// Allocate memory for a, b and c that is accessible to both device and host codes.
hipMallocManaged(&a, sizeof(*a));
hipMallocManaged(&b, sizeof(*b));
hipMallocManaged(&c, sizeof(*c));
// Setup input values.
*a = 1;
*b = 2;
// Launch add() kernel on GPU.
hipLaunchKernelGGL(add, dim3(1), dim3(1), 0, 0, a, b, c);
// Wait for GPU to finish before accessing on host.
hipDeviceSynchronize();
// Query an attribute of the memory range.
hipMemRangeGetAttribute(&attributeValue,
attributeSize,
hipMemRangeAttributeReadMostly,
a,
sizeof(*a));
// Prints the result.
std::cout << *a << " + " << *b << " = " << *c << std::endl;
std::cout << "The queried attribute value is: " << attributeValue << std::endl;
// Cleanup allocated memory.
hipFree(a);
hipFree(b);
hipFree(c);
return 0;
}
Asynchronously attach memory to a stream
----------------------------------------
The ``hipStreamAttachMemAsync`` function would be able to asynchronously attach memory to a stream, which can help concurrent execution when using streams.
Currently, this function is a no-operation (NOP) function on AMD GPUs. It simply returns success after the runtime memory validation passed. This function is necessary on Microsoft Windows, and UMM is not supported on this operating system with AMD GPUs at the moment.
projects/hip/docs/how-to/virtual_memory.rst
+94
Melihat File
@@ -0,0 +1,94 @@
.. meta::
:description: This chapter describes introduces Virtual Memory (VM) and shows
how to use it in AMD HIP.
:keywords: AMD, ROCm, HIP, CUDA, virtual memory, virtual, memory, UM, APU
.. _virtual_memory:
*****************************
Virtual memory management
*****************************
Memory management is important when creating high-performance applications in the HIP ecosystem. Both allocating and copying memory can result in bottlenecks, which can significantly impact performance.
Global memory allocation in HIP uses the C language style allocation function. This works fine for simple cases but can cause problems if your memory needs change. If you need to increase the size of your memory, you must allocate a second larger buffer and copy the data to it before you can free the original buffer. This increases overall memory usage and causes unnecessary ``memcpy`` calls. Another solution is to allocate a larger buffer than you initially need. However, this isn't an efficient way to handle resources and doesn't solve the issue of reallocation when the extra buffer runs out.
Virtual memory management solves these memory management problems. It helps to reduce memory usage and unnecessary ``memcpy`` calls.
.. _memory_allocation_virtual_memory:
Memory allocation
=================
Standard memory allocation uses the ``hipMalloc`` function to allocate a block of memory on the device. However, when using virtual memory, this process is separated into multiple steps using the ``hipMemCreate``, ``hipMemAddressReserve``, ``hipMemMap``, and ``hipMemSetAccess`` functions. This guide explains what these functions do and how you can use them for virtual memory management.
Allocate physical memory
------------------------
The first step is to allocate the physical memory itself with the ``hipMemCreate`` function. This function accepts the size of the buffer, an ``unsigned long long`` variable for the flags, and a ``hipMemAllocationProp`` variable. ``hipMemAllocationProp`` contains the properties of the memory to be allocated, such as where the memory is physically located and what kind of shareable handles are available. If the allocation is successful, the function returns a value of ``hipSuccess``, with ``hipMemGenericAllocationHandle_t`` representing a valid physical memory allocation. The allocated memory size must be aligned with the granularity appropriate for the properties of the allocation. You can use the ``hipMemGetAllocationGranularity`` function to determine the correct granularity.
.. code-block:: cpp
size_t granularity = 0;
hipMemGenericAllocationHandle_t allocHandle;
hipMemAllocationProp prop = {};
prop.type = HIP_MEM_ALLOCATION_TYPE_PINNED;
prop.location.type = HIP_MEM_LOCATION_TYPE_DEVICE;
prop.location.id = currentDev;
hipMemGetAllocationGranularity(&granularity, &prop, HIP_MEM_ALLOC_GRANULARITY_MINIMUM);
padded_size = ROUND_UP(size, granularity);
hipMemCreate(&allocHandle, padded_size, &prop, 0);
Reserve virtual address range
-----------------------------
After you have acquired an allocation of physical memory, you must map it before you can use it. To do so, you need a virtual address to map it to. Mapping means the physical memory allocation is available from the virtual address range it is mapped to. To reserve a virtual memory range, use the ``hipMemAddressReserve`` function. The size of the virtual memory must match the amount of physical memory previously allocated. You can then map the physical memory allocation to the newly-acquired virtual memory address range using the ``hipMemMap`` function.
.. code-block:: cpp
hipMemAddressReserve(&ptr, padded_size, 0, 0, 0);
hipMemMap(ptr, padded_size, 0, allocHandle, 0);
Set memory access
-----------------
Finally, use the ``hipMemSetAccess`` function to enable memory access. It accepts the pointer to the virtual memory, the size, and a ``hipMemAccessDesc`` descriptor as parameters. In a multi-GPU environment, you can map the device memory of one GPU to another. This feature also works with the traditional memory management system, but isn't as scalable as with virtual memory. When memory is allocated with ``hipMalloc``, ``hipDeviceEnablePeerAccess`` is used to enable peer access. This function enables access between two devices, but it means that every call to ``hipMalloc`` takes more time to perform the checks and the mapping between the devices. When using virtual memory management, peer access is enabled by ``hipMemSetAccess``, which provides a finer level of control over what is shared. This has no performance impact on memory allocation and gives you more control over what memory buffers are shared with which devices.
.. code-block:: cpp
hipMemAccessDesc accessDesc = {};
accessDesc.location.type = HIP_MEM_LOCATION_TYPE_DEVICE;
accessDesc.location.id = currentDev;
accessDesc.flags = HIP_MEM_ACCESS_FLAGS_PROT_READWRITE;
hipMemSetAccess(ptr, padded_size, &accessDesc, 1);
At this point the memory is allocated, mapped, and ready for use. You can read and write to it, just like you would a C style memory allocation.
Free virtual memory
-------------------
To free the memory allocated in this manner, use the corresponding free functions. To unmap the memory, use ``hipMemUnmap``. To release the virtual address range, use ``hipMemAddressFree``. Finally, to release the physical memory, use ``hipMemRelease``. A side effect of these functions is the lack of synchronization when memory is released. If you call ``hipFree`` when you have multiple streams running in parallel, it synchronizes the device. This causes worse resource usage and performance.
.. code-block:: cpp
hipMemUnmap(ptr, size);
hipMemRelease(allocHandle);
hipMemAddressFree(ptr, size);
.. _usage_virtual_memory:
Memory usage
============
Dynamically increase allocation size
------------------------------------
The ``hipMemAddressReserve`` function allows you to increase the amount of pre-allocated memory. This function accepts a parameter representing the requested starting address of the virtual memory. This allows you to have a continuous virtual address space without worrying about the underlying physical allocation.
.. code-block:: cpp
hipMemAddressReserve(&new_ptr, (new_size - padded_size), 0, ptr + padded_size, 0);
hipMemMap(new_ptr, (new_size - padded_size), 0, newAllocHandle, 0);
hipMemSetAccess(new_ptr, (new_size - padded_size), &accessDesc, 1);
The code sample above assumes that ``hipMemAddressReserve`` was able to reserve the memory address at the specified location. However, this isn't guaranteed to be true, so you should validate that ``new_ptr`` points to a specific virtual address before using it.
projects/hip/docs/index.md
+22 -9
Melihat File
@@ -11,7 +11,7 @@ For HIP supported AMD GPUs on multiple operating systems, see:
The CUDA enabled NVIDIA GPUs are supported by HIP. For more information, see [GPU Compute Capability](https://developer.nvidia.com/cuda-gpus).
On the AMD ROCm platform, HIP provides header files and runtime library built on top of HIP-Clang compiler in the repository [Common Language Runtime (CLR)](./understand/amd_clr), which contains source codes for AMD's compute languages runtimes as follows,
On the AMD ROCm platform, HIP provides header files and runtime library built on top of HIP-Clang compiler in the repository [Common Language Runtimes (CLR)](./understand/amd_clr), which contains source codes for AMD's compute languages runtimes as follows,
On non-AMD platforms, like NVIDIA, HIP provides header files required to support non-AMD specific back-end implementation in the repository ['hipother'](https://github.com/ROCm/hipother), which translates from the HIP runtime APIs to CUDA runtime APIs.
@@ -30,31 +30,40 @@ On non-AMD platforms, like NVIDIA, HIP provides header files required to support
:::{grid-item-card} Conceptual
* {doc}`./understand/programming_model`
* {doc}`./understand/programming_model_reference`
* {doc}`./understand/hardware_implementation`
* {doc}`./understand/amd_clr`
* {doc}`./understand/texture_fetching`
:::
:::{grid-item-card} How to
* [Programming Manual](./how-to/programming_manual)
* [HIP Porting Guide](./how-to/hip_porting_guide)
* [HIP Porting: Driver API Guide](./how-to/hip_porting_driver_api)
* [Programming manual](./how-to/programming_manual)
* [HIP porting guide](./how-to/hip_porting_guide)
* [HIP porting: driver API guide](./how-to/hip_porting_driver_api)
* {doc}`./how-to/hip_rtc`
* {doc}`./how-to/performance_guidelines`
* [Debugging with HIP](./how-to/debugging)
* {doc}`./how-to/logging`
* [Unified memory](./how-to/unified_memory)
* [Virtual memory](./how-to/virtual_memory)
* {doc}`./how-to/stream_ordered_allocator`
* [Cooperative groups](./how-to/cooperative_groups)
* [HIP graphs](./how-to/hipgraph)
* {doc}`./how-to/faq`
:::
:::{grid-item-card} Reference
* {doc}`/doxygen/html/index`
* [C++ language extensions](./reference/kernel_language)
* [Comparing Syntax for different APIs](./reference/terms)
* [HSA Runtime API for ROCm](./reference/virtual_rocr)
* [HIP runtime API](./reference/hip_runtime_api_reference)
* [Modules](./reference/hip_runtime_api/modules)
* [Global defines, enums, structs and files](./reference/hip_runtime_api/global_defines_enums_structs_files)
* [HSA runtime API for ROCm](./reference/virtual_rocr)
* [C++ language extensions](./reference/cpp_language_extensions)
* [C++ language support](./reference/cpp_language_support)
* [HIP math API](./reference/math_api)
* [Comparing syntax for different APIs](./reference/terms)
* [List of deprecated APIs](./reference/deprecated_api_list)
* [FP8 numbers in HIP](./reference/fp8_numbers)
@@ -62,8 +71,12 @@ On non-AMD platforms, like NVIDIA, HIP provides header files required to support
:::{grid-item-card} Tutorial
* [HIP basic examples](https://github.com/ROCm/rocm-examples/tree/develop/HIP-Basic)
* [HIP examples](https://github.com/ROCm/HIP-Examples)
* [HIP test samples](https://github.com/ROCm/hip-tests/tree/develop/samples)
* [SAXPY tutorial](./tutorial/saxpy)
* [Reduction tutorial](./tutorial/reduction)
* [Cooperative groups tutorial](./tutorial/cooperative_groups_tutorial)
:::
projects/hip/docs/reference/kernel_language.rst → projects/hip/docs/reference/cpp_language_extensions.rst
+35 -1154
Melihat File
File diff ditekan karena terlalu besar Load Diff
projects/hip/docs/reference/cpp_language_support.rst
+171
Melihat File
@@ -0,0 +1,171 @@
.. meta::
:description: This chapter describes the C++ support of the HIP ecosystem
ROCm software.
:keywords: AMD, ROCm, HIP, C++
*******************************************************************************
C++ language support
*******************************************************************************
The ROCm platform enables the power of combined C++ and HIP (Heterogeneous-computing
Interface for Portability) code. This code is compiled with a ``clang`` or ``clang++``
compiler. The official compilers support the HIP platform, or you can use the
``amdclang`` or ``amdclang++`` included in the ROCm installation, which are a wrapper for
the official versions.
The source code is compiled according to the ``C++03``, ``C++11``, ``C++14``, ``C++17``,
and ``C++20`` standards, along with HIP-specific extensions, but is subject to
restrictions. The key restriction is the reduced support of standard library in device
code. This is due to the fact that by default a function is considered to run on host,
except for ``constexpr`` functions, which can run on host and device as well.
.. _language_modern_cpp_support:
Modern C++ support
===============================================================================
C++ is considered a modern programming language as of C++11. This section describes how
HIP supports these new C++ features.
C++11 support
-------------------------------------------------------------------------------
The C++11 standard introduced many new features. These features are supported in HIP host
code, with some notable omissions on the device side. The rule of thumb here is that
``constexpr`` functions work on device, the rest doesn't. This means that some important
functionality like ``std::function`` is missing on the device, but unfortunately the
standard library wasn't designed with HIP in mind, which means that the support is in a
state of "works as-is".
Certain features have restrictions and clarifications. For example, any functions using
the ``constexpr`` qualifier or the new ``initializer lists``, ``std::move`` or
``std::forward`` features are implicitly considered to have the ``__host__`` and
``__device__`` execution space specifier. Also, ``constexpr`` variables that are static
members or namespace scoped can be used from both host and device, but only for read
access. Dereferencing a static ``constexpr`` outside its specified execution space causes
an error.
Lambdas are supported, but there are some extensions and restrictions on their usage. For
more information, see the `Extended lambdas`_ section below.
C++14 support
-------------------------------------------------------------------------------
The C++14 language features are supported.
C++17 support
-------------------------------------------------------------------------------
All C++17 language features are supported.
C++20 support
-------------------------------------------------------------------------------
All C++20 language features are supported, but extensions and restrictions apply. C++20
introduced coroutines and modules, which fundamentally changed how programs are written.
HIP doesn't support these features. However, ``consteval`` functions can be called from
host and device, even if specified for host use only.
The three-way comparison operator (spaceship operator ``<=>``) works with host and device
code.
.. _language_restrictions:
Extensions and restrictions
===============================================================================
In addition to the deviations from the standard, there are some general extensions and
restrictions to consider.
Global functions
-------------------------------------------------------------------------------
Functions that serve as an entry point for device execution are called kernels and are
specified with the ``__global__`` qualifier. To call a kernel function, use the triple
chevron operator: ``<<< >>>``. Kernel functions must have a ``void`` return type. These
functions can't:
* have a ``constexpr`` specifier
* have a parameter of type ``std::initializer_list`` or ``va_list``
* use an rvalue reference as a parameter.
* use parameters having different sizes in host and device code, e.g. long double arguments, or structs containing long double members.
* use struct-type arguments which have different layout in host and device code.
Kernels can have variadic template parameters, but only one parameter pack, which must be
the last item in the template parameter list.
Device space memory specifiers
-------------------------------------------------------------------------------
HIP includes device space memory specifiers to indicate whether a variable is allocated
in host or device memory and how its memory should be allocated. HIP supports the
``__device__``, ``__shared__``, ``__managed__``, and ``__constant__`` specifiers.
The ``__device__`` and ``__constant__`` specifiers define global variables, which are
allocated within global memory on the HIP devices. The only difference is that
``__constant__`` variables can't be changed after allocation. The ``__shared__``
specifier allocates the variable within shared memory, which is available for all threads
in a block.
The ``__managed__`` variable specifier creates global variables that are initially
undefined and unaddressed within the global symbol table. The HIP runtime allocates
managed memory and defines the symbol when it loads the device binary. A managed variable
can be accessed in both device and host code.
It's important to know where a variable is stored because it is only available from
certain locations. Generally, variables allocated in the host memory are not accessible
from the device code, while variables allocated in the device memory are not directly
accessible from the host code. Dereferencing a pointer to device memory on the host
results in a segmentation fault. Accessing device variables in host code should be done
through kernel execution or HIP functions like ``hipMemCpyToSymbol``.
Exception handling
-------------------------------------------------------------------------------
An important difference between the host and device code is exception handling. In device
code, this control flow isn't available due to the hardware architecture. The device
code must use return codes to handle errors.
Kernel parameters
-------------------------------------------------------------------------------
There are some restrictions on kernel function parameters. They cannot be passed by
reference, because these functions are called from the host but run on the device. Also,
a variable number of arguments is not allowed.
Classes
-------------------------------------------------------------------------------
Classes work on both the host and device side, but there are some constraints. The
``static`` member functions can't be ``__global__``. ``Virtual`` member functions work,
but a ``virtual`` function must not be called from the host if the parent object was
created on the device, or the other way around, because this behavior is undefined.
Another minor restriction is that ``__device__`` variables, that are global scoped must
have trivial constructors.
Polymorphic function wrappers
-------------------------------------------------------------------------------
HIP doesn't support the polymorphic function wrapper ``std::function``, which was
introduced in C++11.
Extended lambdas
-------------------------------------------------------------------------------
HIP supports Lambdas, which by default work as expected.
Lambdas have implicit host device attributes. This means that they can be executed by
both host and device code, and works the way you would expect. To make a lambda callable
only by host or device code, users can add ``__host__`` or ``__device__`` attribute. The
only restriction is that host variables can only be accessed through copy on the device.
Accessing through reference will cause undefined behavior.
Inline namespaces
-------------------------------------------------------------------------------
Inline namespaces are supported, but with a few exceptions. The following entities can't
be declared in namespace scope within an inline unnamed namespace:
* ``__managed__``, ``__device__``, ``__shared__`` and ``__constant__`` variables
* ``__global__`` function and function templates
* variables with surface or texture type
projects/hip/docs/reference/fp8_numbers.rst
+1 -1
Melihat File
@@ -28,7 +28,7 @@ There are two formats of FP8 numbers, E4M3 and E5M2.
HIP Header
==========
HIP header defined the FP8 ocp/fnuz numbers `here <https://github.com/ROCm/clr/blob/develop/hipamd/include/hip/amd_detail/amd_hip_fp8.h>`_.
The `HIP header <https://github.com/ROCm/clr/blob/develop/hipamd/include/hip/amd_detail/amd_hip_fp8.h>`_ defines the FP8 ocp/fnuz numbers.
Supported Devices
=================
projects/hip/docs/reference/hip_runtime_api/global_defines_enums_structs_files.rst
+15
Melihat File
@@ -0,0 +1,15 @@
.. meta::
:description: The global defines, enum, structs and files reference page.
.. _global_defines_enums_structs_files_reference:
*******************************************************************************
Global defines, enums, structs and files
*******************************************************************************
The structs, define macros, enums and files in the HIP runtime API.
* :ref:`global_enum_defines_reference`
* :ref:`driver_types_reference`
* :doc:`hip:doxygen/html/annotated`
* :doc:`hip:doxygen/html/files`
projects/hip/docs/reference/hip_runtime_api/global_defines_enums_structs_files/driver_types.rst
+12
Melihat File
@@ -0,0 +1,12 @@
.. meta::
:description: The driver types reference page.
:keywords: AMD, ROCm, HIP, CUDA, driver types
.. _driver_types_reference:
*******************************************************************************
Driver types
*******************************************************************************
.. doxygengroup:: DriverTypes
:content-only:
projects/hip/docs/reference/hip_runtime_api/global_defines_enums_structs_files/global_enum_and_defines.rst
+12
Melihat File
@@ -0,0 +1,12 @@
.. meta::
:description: The global enum and defines reference page.
:keywords: AMD, ROCm, HIP, CUDA, global enum, defines
.. _global_enum_defines_reference:
*******************************************************************************
Global enum and defines
*******************************************************************************
.. doxygengroup:: GlobalDefs
:content-only:
projects/hip/docs/reference/hip_runtime_api/modules.rst
+41
Melihat File
@@ -0,0 +1,41 @@
.. meta::
:description: The HIP runtime API modules reference page.
:keywords: AMD, ROCm, HIP, CUDA, HIP runtime API modules, modules
.. _modules_reference:
*******************************************************************************
Modules
*******************************************************************************
The API is organized into modules based on functionality.
* :ref:`initialization_version_reference`
* :ref:`device_management_reference`
* :ref:`execution_control_reference`
* :ref:`error_handling_reference`
* :ref:`stream_management_reference`
* :ref:`stream_memory_operations_reference`
* :ref:`event_management_reference`
* :ref:`memory_management_reference`
* :ref:`memory_management_deprecated_reference`
* :ref:`external_resource_interoperability_reference`
* :ref:`stream_ordered_memory_allocator_reference`
* :ref:`unified_memory_reference`
* :ref:`virtual_memory_reference`
* :ref:`texture_management_reference`
* :ref:`texture_management_deprecated_reference`
* :ref:`surface_object_reference`
* :ref:`peer_to_peer_device_memory_access_reference`
* :ref:`context_management_reference`
* :ref:`module_management_reference`
* :ref:`occupancy_reference`
* :ref:`profiler_control_reference`
* :ref:`launch_api_reference`
* :ref:`runtime_compilation_reference`
* :ref:`callback_activity_apis_reference`
* :ref:`graph_management_reference`
* :ref:`opengl_interoperability_reference`
* :ref:`cooperative_groups_reference`
projects/hip/docs/reference/hip_runtime_api/modules/callback_activity_apis.rst
+12
Melihat File
@@ -0,0 +1,12 @@
.. meta::
:description: The callback activity APIs reference page.
:keywords: AMD, ROCm, HIP, CUDA, callback activity APIs, callback activity
.. _callback_activity_apis_reference:
*******************************************************************************
Callback activity APIs
*******************************************************************************
.. doxygengroup:: Callback
:content-only:
projects/hip/docs/reference/hip_runtime_api/modules/context_management.rst
+12
Melihat File
@@ -0,0 +1,12 @@
.. meta::
:description: The context management reference page.
:keywords: AMD, ROCm, HIP, CUDA, context management, context
.. _context_management_reference:
*******************************************************************************
Context management [deprecated]
*******************************************************************************
.. doxygengroup:: Context
:content-only:
projects/hip/docs/reference/hip_runtime_api/modules/cooperative_groups_reference.rst
+70
Melihat File
@@ -0,0 +1,70 @@
.. meta::
:description: This chapter lists types and device API wrappers related to the
Cooperative Group feature. Programmers can directly use these
API features in their kernels.
:keywords: AMD, ROCm, HIP, cooperative groups
.. _cooperative_groups_reference:
*******************************************************************************
Cooperative groups
*******************************************************************************
Cooperative kernel launches
===========================
The following host-side functions are used for cooperative kernel launches.
.. doxygengroup:: ModuleCooperativeG
:content-only:
Cooperative groups classes
==========================
The following cooperative groups classes can be used on the device side.
.. _thread_group_ref:
.. doxygenclass:: cooperative_groups::thread_group
:members:
.. _thread_block_ref:
.. doxygenclass:: cooperative_groups::thread_block
:members:
.. _grid_group_ref:
.. doxygenclass:: cooperative_groups::grid_group
:members:
.. _multi_grid_group_ref:
.. doxygenclass:: cooperative_groups::multi_grid_group
:members:
.. _thread_block_tile_ref:
.. doxygenclass:: cooperative_groups::thread_block_tile
:members:
.. _coalesced_group_ref:
.. doxygenclass:: cooperative_groups::coalesced_group
:members:
Cooperative groups construct functions
======================================
The following functions are used to construct different group-type instances on the device side.
.. doxygengroup:: CooperativeGConstruct
:content-only:
Cooperative groups exposed API functions
========================================
The following functions are the exposed API for different group-type instances on the device side.
.. doxygengroup:: CooperativeGAPI
:content-only:
projects/hip/docs/reference/hip_runtime_api/modules/device_management.rst
+12
Melihat File
@@ -0,0 +1,12 @@
.. meta::
:description: The device management reference page.
:keywords: AMD, ROCm, HIP, CUDA, device management, device
.. _device_management_reference:
*******************************************************************************
Device management
*******************************************************************************
.. doxygengroup:: Device
:content-only:
projects/hip/docs/reference/hip_runtime_api/modules/error_handling.rst
+12
Melihat File
@@ -0,0 +1,12 @@
.. meta::
:description: The error handling reference page.
:keywords: AMD, ROCm, HIP, CUDA, error handling, error
.. _error_handling_reference:
*******************************************************************************
Error handling
*******************************************************************************
.. doxygengroup:: Error
:content-only:
projects/hip/docs/reference/hip_runtime_api/modules/event_management.rst
+12
Melihat File
@@ -0,0 +1,12 @@
.. meta::
:description: The event management reference page.
:keywords: AMD, ROCm, HIP, CUDA, event management, event
.. _event_management_reference:
*******************************************************************************
Event management
*******************************************************************************
.. doxygengroup:: Event
:content-only:
projects/hip/docs/reference/hip_runtime_api/modules/execution_control.rst
+12
Melihat File
@@ -0,0 +1,12 @@
.. meta::
:description: The execution control reference page.
:keywords: AMD, ROCm, HIP, CUDA, execution control, execution
.. _execution_control_reference:
*******************************************************************************
Execution control
*******************************************************************************
.. doxygengroup:: Execution
:content-only:
projects/hip/docs/reference/hip_runtime_api/modules/graph_management.rst
+12
Melihat File
@@ -0,0 +1,12 @@
.. meta::
:description: The graph management reference page.
:keywords: AMD, ROCm, HIP, CUDA, graph management, graph
.. _graph_management_reference:
*******************************************************************************
Graph management
*******************************************************************************
.. doxygengroup:: Graph
:content-only:
projects/hip/docs/reference/hip_runtime_api/modules/initialization_and_version.rst
+12
Melihat File
@@ -0,0 +1,12 @@
.. meta::
:description: The initialization and version reference page.
:keywords: AMD, ROCm, HIP, CUDA, initialization, version
.. _initialization_version_reference:
*******************************************************************************
Initialization and version
*******************************************************************************
.. doxygengroup:: Driver
:content-only:
projects/hip/docs/reference/hip_runtime_api/modules/launch_api.rst
+12
Melihat File
@@ -0,0 +1,12 @@
.. meta::
:description: The launch API reference page.
:keywords: AMD, ROCm, HIP, CUDA, launch API, triple-chevron
.. _launch_api_reference:
*******************************************************************************
Launch API
*******************************************************************************
.. doxygengroup:: Clang
:content-only:
projects/hip/docs/reference/hip_runtime_api/modules/memory_management.rst
+12
Melihat File
@@ -0,0 +1,12 @@
.. meta::
:description: The memory management reference page.
:keywords: AMD, ROCm, HIP, CUDA, memory management, memory
.. _memory_management_reference:
*******************************************************************************
Memory management
*******************************************************************************
.. doxygengroup:: Memory
:content-only:
projects/hip/docs/reference/hip_runtime_api/modules/memory_management/external_resource_interoperability.rst
+12
Melihat File
@@ -0,0 +1,12 @@
.. meta::
:description: The external resource interoperability reference page.
:keywords: AMD, ROCm, HIP, CUDA, external resource interoperability
.. _external_resource_interoperability_reference:
*******************************************************************************
External resource interoperability
*******************************************************************************
.. doxygengroup:: External
:content-only:
projects/hip/docs/reference/hip_runtime_api/modules/memory_management/memory_management_deprecated.rst
+11
Melihat File
@@ -0,0 +1,11 @@
.. meta::
:description: The deprecated memory management reference page.
.. _memory_management_deprecated_reference:
*******************************************************************************
Memory management (deprecated)
*******************************************************************************
.. doxygengroup:: MemoryD
:content-only:
projects/hip/docs/reference/hip_runtime_api/modules/memory_management/stream_ordered_memory_allocator.rst
+12
Melihat File
@@ -0,0 +1,12 @@
.. meta::
:description: The stream ordered memory allocator reference page.
:keywords: AMD, ROCm, HIP, CUDA, stream ordered memory allocator
.. _stream_ordered_memory_allocator_reference:
*******************************************************************************
Stream ordered memory allocator
*******************************************************************************
.. doxygengroup:: StreamO
:content-only:
projects/hip/docs/reference/hip_runtime_api/modules/memory_management/surface_object.rst
+15
Melihat File
@@ -0,0 +1,15 @@
.. meta::
:description: The surface object reference page.
:keywords: AMD, ROCm, HIP, CUDA, surface object, surface
.. _surface_object_reference:
*******************************************************************************
Surface object
*******************************************************************************
.. doxygengroup:: Surface
:content-only:
.. doxygengroup:: SurfaceAPI
:content-only:
projects/hip/docs/reference/hip_runtime_api/modules/memory_management/texture_management.rst
+12
Melihat File
@@ -0,0 +1,12 @@
.. meta::
:description: The texture management reference page.
:keywords: AMD, ROCm, HIP, CUDA, texture management, texture
.. _texture_management_reference:
*******************************************************************************
Texture management
*******************************************************************************
.. doxygengroup:: Texture
:content-only:
projects/hip/docs/reference/hip_runtime_api/modules/memory_management/texture_management_deprecated.rst
+12
Melihat File
@@ -0,0 +1,12 @@
.. meta::
:description: The deprecated texture management reference page.
:keywords: AMD, ROCm, HIP, CUDA, deprecated texture management
.. _texture_management_deprecated_reference:
*******************************************************************************
Texture management (deprecated)
*******************************************************************************
.. doxygengroup:: TextureD
:content-only:
projects/hip/docs/reference/hip_runtime_api/modules/memory_management/unified_memory_reference.rst
+12
Melihat File
@@ -0,0 +1,12 @@
.. meta::
:description: The managed memory reference page.
:keywords: AMD, ROCm, HIP, CUDA, unified memory, unified, memory, UM, APU
.. _unified_memory_reference:
*******************************************************************************
Managed memory
*******************************************************************************
.. doxygengroup:: MemoryM
:content-only:
projects/hip/docs/reference/hip_runtime_api/modules/memory_management/virtual_memory_reference.rst
+12
Melihat File
@@ -0,0 +1,12 @@
.. meta::
:description: The virtual memory (VM) management reference page.
:keywords: AMD, ROCm, HIP, CUDA, virtual memory, virtual, memory, VM
.. _virtual_memory_reference:
*******************************************************************************
Virtual memory management
*******************************************************************************
.. doxygengroup:: Virtual
:content-only:
projects/hip/docs/reference/hip_runtime_api/modules/module_management.rst
+12
Melihat File
@@ -0,0 +1,12 @@
.. meta::
:description: The module management reference page.
:keywords: AMD, ROCm, HIP, CUDA, module management, module
.. _module_management_reference:
*******************************************************************************
Module management
*******************************************************************************
.. doxygengroup:: Module
:content-only:
projects/hip/docs/reference/hip_runtime_api/modules/occupancy.rst
+12
Melihat File
@@ -0,0 +1,12 @@
.. meta::
:description: The occupancy reference page.
:keywords: AMD, ROCm, HIP, CUDA, occupancy
.. _occupancy_reference:
*******************************************************************************
Occupancy
*******************************************************************************
.. doxygengroup:: Occupancy
:content-only:
projects/hip/docs/reference/hip_runtime_api/modules/opengl_interoperability.rst
+12
Melihat File
@@ -0,0 +1,12 @@
.. meta::
:description: The OpenGL interoperability reference page.
:keywords: AMD, ROCm, HIP, CUDA, OpenGL interoperability, OpenGL interop
.. _opengl_interoperability_reference:
*******************************************************************************
OpenGL interoperability
*******************************************************************************
.. doxygengroup:: GL
:content-only:
projects/hip/docs/reference/hip_runtime_api/modules/peer_to_peer_device_memory_access.rst
+12
Melihat File
@@ -0,0 +1,12 @@
.. meta::
:description: The peer to peer device memory access reference page.
:keywords: AMD, ROCm, HIP, CUDA, peer to peer device memory access, peer to peer
.. _peer_to_peer_device_memory_access_reference:
*******************************************************************************
Peer to peer device memory access
*******************************************************************************
.. doxygengroup:: PeerToPeer
:content-only:
projects/hip/docs/reference/hip_runtime_api/modules/profiler_control.rst
+12
Melihat File
@@ -0,0 +1,12 @@
.. meta::
:description: The profiler control reference page.
:keywords: AMD, ROCm, HIP, CUDA, profiler control, profiler
.. _profiler_control_reference:
*******************************************************************************
Profiler control
*******************************************************************************
.. doxygengroup:: Profiler
:content-only:
projects/hip/docs/reference/hip_runtime_api/modules/runtime_compilation.rst
+12
Melihat File
@@ -0,0 +1,12 @@
.. meta::
:description: The runtime compilation reference page.
:keywords: AMD, ROCm, HIP, CUDA, runtime compilation
.. _runtime_compilation_reference:
*******************************************************************************
Runtime compilation
*******************************************************************************
.. doxygengroup:: Runtime
:content-only:
projects/hip/docs/reference/hip_runtime_api/modules/stream_management.rst
+12
Melihat File
@@ -0,0 +1,12 @@
.. meta::
:description: The stream management reference page.
:keywords: AMD, ROCm, HIP, CUDA, stream management, stream
.. _stream_management_reference:
*******************************************************************************
Stream management
*******************************************************************************
.. doxygengroup:: Stream
:content-only:
projects/hip/docs/reference/hip_runtime_api/modules/stream_memory_operations.rst
+12
Melihat File
@@ -0,0 +1,12 @@
.. meta::
:description: The stream memory operations reference page.
:keywords: AMD, ROCm, HIP, CUDA, stream memory operations
.. _stream_memory_operations_reference:
*******************************************************************************
Stream memory operations
*******************************************************************************
.. doxygengroup:: StreamM
:content-only:
projects/hip/docs/reference/hip_runtime_api_reference.rst
+14
Melihat File
@@ -0,0 +1,14 @@
.. meta::
:description: HIP runtime API reference page
:keywords: AMD, ROCm, HIP, CUDA, HIP runtime API, HIP runtime
.. _runtime_api_reference:
********************************************************************************
HIP runtime API
********************************************************************************
The HIP Runtime API reference:
* :ref:`modules_reference`
* :ref:`global_defines_enums_structs_files_reference`
projects/hip/docs/reference/math_api.rst
+1121
Melihat File
File diff ditekan karena terlalu besar Load Diff
projects/hip/docs/reference/terms.md
+1 -1
Melihat File
@@ -1,4 +1,4 @@
# Table Comparing Syntax for Different Compute APIs
# Table comparing syntax for different compute APIs
|Term|CUDA|HIP|OpenCL|
|---|---|---|---|
projects/hip/docs/reference/virtual_rocr.rst
+1 -1
Melihat File
@@ -5,7 +5,7 @@
:keywords: AMD, ROCm, HIP, HSA, ROCR runtime, virtual memory management
*******************************************************************************
HSA Runtime API for ROCm
HSA runtime API for ROCm
*******************************************************************************
The following functions are located in the https://github.com/ROCm/ROCR-Runtime repository.
projects/hip/docs/sphinx/_toc.yml.in
+64 -6
Melihat File
@@ -16,9 +16,10 @@ subtrees:
- caption: Conceptual
entries:
- file: understand/programming_model
- file: understand/programming_model_reference
- file: understand/hardware_implementation
- file: understand/amd_clr
- file: understand/texture_fetching
title: Texture fetching
- caption: How to
entries:
@@ -29,16 +30,68 @@ subtrees:
- file: how-to/performance_guidelines
- file: how-to/debugging
- file: how-to/logging
- file: how-to/cooperative_groups
- file: how-to/unified_memory
title: Unified memory
- file: how-to/virtual_memory
title: Virtual memory
- file: how-to/stream_ordered_allocator
- file: how-to/hipgraph
title: HIP graphs
- file: how-to/faq
- caption: Reference
entries:
- file: doxygen/html/index
- file: reference/kernel_language
title: C++ language extensions
- file: reference/terms
title: Comparing Syntax for different APIs
- file: reference/hip_runtime_api_reference
subtrees:
- entries:
- file: reference/hip_runtime_api/modules
subtrees:
- entries:
- file: reference/hip_runtime_api/modules/initialization_and_version
- file: reference/hip_runtime_api/modules/device_management
- file: reference/hip_runtime_api/modules/execution_control
- file: reference/hip_runtime_api/modules/error_handling
- file: reference/hip_runtime_api/modules/stream_management
- file: reference/hip_runtime_api/modules/stream_memory_operations
- file: reference/hip_runtime_api/modules/event_management
- file: reference/hip_runtime_api/modules/memory_management
subtrees:
- entries:
- file: reference/hip_runtime_api/modules/memory_management/memory_management_deprecated
- file: reference/hip_runtime_api/modules/memory_management/external_resource_interoperability
- file: reference/hip_runtime_api/modules/memory_management/stream_ordered_memory_allocator
- file: reference/hip_runtime_api/modules/memory_management/unified_memory_reference
- file: reference/hip_runtime_api/modules/memory_management/virtual_memory_reference
- file: reference/hip_runtime_api/modules/memory_management/texture_management
- file: reference/hip_runtime_api/modules/memory_management/texture_management_deprecated
- file: reference/hip_runtime_api/modules/memory_management/surface_object
- file: reference/hip_runtime_api/modules/peer_to_peer_device_memory_access
- file: reference/hip_runtime_api/modules/context_management
- file: reference/hip_runtime_api/modules/module_management
- file: reference/hip_runtime_api/modules/occupancy
- file: reference/hip_runtime_api/modules/profiler_control
- file: reference/hip_runtime_api/modules/launch_api
- file: reference/hip_runtime_api/modules/runtime_compilation
- file: reference/hip_runtime_api/modules/callback_activity_apis
- file: reference/hip_runtime_api/modules/graph_management
- file: reference/hip_runtime_api/modules/opengl_interoperability
- file: reference/hip_runtime_api/modules/cooperative_groups_reference
- file: reference/hip_runtime_api/global_defines_enums_structs_files
subtrees:
- entries:
- file: reference/hip_runtime_api/global_defines_enums_structs_files/global_enum_and_defines
- file: reference/hip_runtime_api/global_defines_enums_structs_files/driver_types
- file: doxygen/html/annotated
- file: doxygen/html/files
- file: reference/virtual_rocr
- file: reference/cpp_language_extensions
title: C++ language extensions
- file: reference/cpp_language_support
title: C++ language support
- file: reference/math_api
- file: reference/terms
title: Comparing syntax for different APIs
- file: reference/deprecated_api_list
title: List of deprecated APIs
- file: reference/fp8_numbers
@@ -46,10 +99,15 @@ subtrees:
- caption: Tutorials
entries:
- url: https://github.com/ROCm/rocm-examples/tree/develop/HIP-Basic
title: HIP basic examples
- url: https://github.com/ROCm/HIP-Examples
title: HIP examples
- url: https://github.com/ROCm/hip-tests/tree/develop/samples
title: HIP test samples
- file: tutorial/saxpy
- file: tutorial/reduction
- file: tutorial/cooperative_groups_tutorial
- caption: About
entries:
projects/hip/docs/sphinx/requirements.in
+1 -1
Melihat File
@@ -1,2 +1,2 @@
rocm-docs-core[api_reference]==1.1.1
rocm-docs-core[api_reference]==1.7.2
sphinxcontrib.doxylink
projects/hip/docs/sphinx/requirements.txt
+37 -35
Melihat File
@@ -4,11 +4,11 @@
#
# pip-compile requirements.in
#
accessible-pygments==0.0.4
accessible-pygments==0.0.5
# via pydata-sphinx-theme
alabaster==0.7.16
alabaster==1.0.0
# via sphinx
babel==2.14.0
babel==2.16.0
# via
# pydata-sphinx-theme
# sphinx
@@ -16,9 +16,9 @@ beautifulsoup4==4.12.3
# via pydata-sphinx-theme
breathe==4.35.0
# via rocm-docs-core
certifi==2024.2.2
certifi==2024.8.30
# via requests
cffi==1.16.0
cffi==1.17.1
# via
# cryptography
# pynacl
@@ -31,7 +31,7 @@ click==8.1.7
# sphinx-external-toc
click-log==0.4.0
# via doxysphinx
cryptography==42.0.5
cryptography==43.0.1
# via pyjwt
deprecated==1.2.14
# via pygithub
@@ -41,19 +41,19 @@ docutils==0.21.2
# myst-parser
# pydata-sphinx-theme
# sphinx
doxysphinx==3.3.7
doxysphinx==3.3.10
# via rocm-docs-core
fastjsonschema==2.19.1
fastjsonschema==2.20.0
# via rocm-docs-core
gitdb==4.0.11
# via gitpython
gitpython==3.1.43
# via rocm-docs-core
idna==3.7
idna==3.8
# via requests
imagesize==1.4.1
# via sphinx
jinja2==3.1.3
jinja2==3.1.4
# via
# myst-parser
# sphinx
@@ -67,27 +67,29 @@ markdown-it-py==3.0.0
# myst-parser
markupsafe==2.1.5
# via jinja2
mdit-py-plugins==0.4.0
mdit-py-plugins==0.4.1
# via myst-parser
mdurl==0.1.2
# via markdown-it-py
mpire==2.10.1
mpire==2.10.2
# via doxysphinx
myst-parser==3.0.0
myst-parser==4.0.0
# via rocm-docs-core
packaging==24.0
numpy==1.26.4
# via doxysphinx
packaging==24.1
# via
# pydata-sphinx-theme
# sphinx
pycparser==2.22
# via cffi
pydata-sphinx-theme==0.15.2
pydata-sphinx-theme==0.15.4
# via
# rocm-docs-core
# sphinx-book-theme
pygithub==2.3.0
pygithub==2.4.0
# via rocm-docs-core
pygments==2.17.2
pygments==2.18.0
# via
# accessible-pygments
# mpire
@@ -95,26 +97,26 @@ pygments==2.17.2
# sphinx
pyjson5==1.6.6
# via doxysphinx
pyjwt[crypto]==2.8.0
pyjwt[crypto]==2.9.0
# via pygithub
pynacl==1.5.0
# via pygithub
pyparsing==3.1.2
pyparsing==3.1.4
# via
# doxysphinx
# sphinxcontrib-doxylink
python-dateutil==2.9.0.post0
# via sphinxcontrib-doxylink
pyyaml==6.0.1
pyyaml==6.0.2
# via
# myst-parser
# rocm-docs-core
# sphinx-external-toc
requests==2.31.0
requests==2.32.3
# via
# pygithub
# sphinx
rocm-docs-core[api-reference]==1.1.1
rocm-docs-core[api-reference]==1.7.2
# via -r requirements.in
six==1.16.0
# via python-dateutil
@@ -122,9 +124,9 @@ smmap==5.0.1
# via gitdb
snowballstemmer==2.2.0
# via sphinx
soupsieve==2.5
soupsieve==2.6
# via beautifulsoup4
sphinx==7.3.7
sphinx==8.0.2
# via
# breathe
# myst-parser
@@ -136,39 +138,39 @@ sphinx==7.3.7
# sphinx-external-toc
# sphinx-notfound-page
# sphinxcontrib-doxylink
sphinx-book-theme==1.1.2
sphinx-book-theme==1.1.3
# via rocm-docs-core
sphinx-copybutton==0.5.2
# via rocm-docs-core
sphinx-design==0.5.0
sphinx-design==0.6.1
# via rocm-docs-core
sphinx-external-toc==1.0.1
# via rocm-docs-core
sphinx-notfound-page==1.0.0
sphinx-notfound-page==1.0.4
# via rocm-docs-core
sphinxcontrib-applehelp==1.0.8
sphinxcontrib-applehelp==2.0.0
# via sphinx
sphinxcontrib-devhelp==1.0.6
sphinxcontrib-devhelp==2.0.0
# via sphinx
sphinxcontrib-doxylink==1.12.3
# via -r requirements.in
sphinxcontrib-htmlhelp==2.0.5
sphinxcontrib-htmlhelp==2.1.0
# via sphinx
sphinxcontrib-jsmath==1.0.1
# via sphinx
sphinxcontrib-qthelp==1.0.7
sphinxcontrib-qthelp==2.0.0
# via sphinx
sphinxcontrib-serializinghtml==1.1.10
sphinxcontrib-serializinghtml==2.0.0
# via sphinx
tomli==2.0.1
# via sphinx
tqdm==4.66.2
tqdm==4.66.5
# via mpire
typing-extensions==4.11.0
typing-extensions==4.12.2
# via
# pydata-sphinx-theme
# pygithub
urllib3==2.2.1
urllib3==2.2.2
# via
# pygithub
# requests
projects/hip/docs/tutorial/cooperative_groups_tutorial.rst
+240
Melihat File
@@ -0,0 +1,240 @@
.. meta::
:description: HIP cooperative groups tutorial
:keywords: AMD, ROCm, HIP, cooperative groups, tutorial
*******************************************************************************
Cooperative groups
*******************************************************************************
This tutorial demonstrates the basic concepts of cooperative groups in the HIP (Heterogeneous-computing Interface for Portability) programming model and the most essential tooling supporting it. This topic also reviews the commonalities of heterogeneous APIs. Familiarity with the C/C++ compilation model and the language is assumed.
Prerequisites
=============
To follow this tutorial, you'll need properly installed drivers and a HIP compiler toolchain to compile your code. Because ROCm HIP supports compiling and running on Linux and Microsoft Windows with AMD and NVIDIA GPUs, review the HIP development package installation before starting this tutorial. For more information, see :doc:`/install/install`.
Simple HIP Code
===============
To become familiar with heterogeneous programming, review the :doc:`SAXPY tutorial <saxpy>` and the first HIP code subsection. Compiling is also described in that tutorial.
Tiled partition
===============
You can use tiled partition to calculate the sum of ``partition_size`` length sequences and the sum of ``result_size``/ ``BlockSize`` length sequences. The host-side reference implementation is the following:
.. code-block:: cpp
// Host-side function to perform the same reductions as executed on the GPU
std::vector<unsigned int> ref_reduced(const unsigned int partition_size,
std::vector<unsigned int> input)
{
const unsigned int input_size = input.size();
const unsigned int result_size = input_size / partition_size;
std::vector<unsigned int> result(result_size);
for(unsigned int i = 0; i < result_size; i++)
{
unsigned int partition_result = 0;
for(unsigned int j = 0; j < partition_size; j++)
{
partition_result += input[partition_size * i + j];
}
result[i] = partition_result;
}
return result;
}
Device-side code
----------------
To calculate the sum of the sets of numbers, the tutorial uses the shared memory-based reduction on the device side. The warp level intrinsics usage is not covered in this tutorial, unlike in the :doc:`reduction tutorial. <reduction>` ``x`` input variable is a shared pointer, which needs to be synchronized after every value change. The ``thread_group`` input parameter can be ``thread_block_tile`` or ``thread_block`` because the ``thread_group`` is the parent class of these types. The ``val`` are the numbers to calculate the sum of. The returned results of this function return the final results of the reduction on thread ID 0 of the ``thread_group``, and for every other thread, the function results are 0.
.. code-block:: cuda
/// \brief Summation of `unsigned int val`'s in `thread_group g` using shared memory `x`
__device__ unsigned int reduce_sum(thread_group g, unsigned int* x, unsigned int val)
{
// Rank of this thread in the group
const unsigned int group_thread_id = g.thread_rank();
// We start with half the group size as active threads
// Every iteration the number of active threads halves, until we processed all values
for(unsigned int i = g.size() / 2; i > 0; i /= 2)
{
// Store value for this thread in a shared, temporary array
x[group_thread_id] = val;
// Synchronize all threads in the group
g.sync();
// If our thread is still active, sum with its counterpart in the other half
if(group_thread_id < i)
{
val += x[group_thread_id + i];
}
// Synchronize all threads in the group
g.sync();
}
// Only the first thread returns a valid value
if(g.thread_rank() == 0)
return val;
else
return 0;
}
The ``reduce_sum`` device function is reused to calculate the block and custom
partition sum of the input numbers. The kernel has three sections:
1. Initialization of the reduction function variables.
2. The reduction of thread block and store the results in global memory.
3. The reduction of custom partition and store the results in global memory.
1. Initialization of the reduction function variables
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
In this code section, the shared memory is declared, the thread_block_group and
custom_partition are defined, and the input variables are loaded from global
memory.
.. code-block:: cuda
// threadBlockGroup consists of all threads in the block
thread_block thread_block_group = this_thread_block();
// Workspace array in shared memory required for reduction
__shared__ unsigned int workspace[2048];
unsigned int output;
// Input to reduce
const unsigned int input = d_vector[thread_block_group.thread_rank()];
// ...
// Every custom_partition group consists of 16 threads
thread_block_tile<PartitionSize> custom_partition
= tiled_partition<PartitionSize>(thread_block_group);
2. The reduction of thread block
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
In this code section, the sum is calculated on ``thread_block_group`` level, then the results are stored in global memory.
.. code-block:: cuda
// Perform reduction
output = reduce_sum(thread_block_group, workspace, input);
// Only the first thread returns a valid value
if(thread_block_group.thread_rank() == 0)
{
d_block_reduced_vector[0] = output;
}
3. The reduction of custom partition
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
In this code section, the sum is calculated on the custom partition level, then the results are stored in global memory. The custom partition is a partial block of the thread block, it means the reduction calculates on a shorter sequence of input numbers than at the ``thread_block_group`` case.
.. code-block:: cuda
// Perform reduction
output = reduce_sum(custom_partition, &workspace[group_offset], input);
// Only the first thread in each partition returns a valid value
if(custom_partition.thread_rank() == 0)
{
const unsigned int partition_id = thread_block_group.thread_rank() / PartitionSize;
d_partition_reduced_vector[partition_id] = output;
}
Host-side code
--------------
On the host-side, the following steps are done in the example:
1. Confirm the cooperative group support on AMD GPUs.
2. Initialize the cooperative group configuration.
3. Allocate and copy input to global memory.
4. Launch the cooperative kernel.
5. Save the results from global memory.
6. Free the global memory.
Only the first, second and fourth steps are important from the cooperative groups aspect, that's why those steps are detailed further.
1. Confirm the cooperative group support on AMD GPUs
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Not all AMD GPUs support cooperative groups. You can confirm support with the following code:
.. code-block:: cpp
#ifdef __HIP_PLATFORM_AMD__
int device = 0;
int supports_coop_launch = 0;
// Check support
// Use hipDeviceAttributeCooperativeMultiDeviceLaunch when launching across multiple devices
HIP_CHECK(hipGetDevice(&device));
HIP_CHECK(
hipDeviceGetAttribute(&supports_coop_launch, hipDeviceAttributeCooperativeLaunch, device));
if(!supports_coop_launch)
{
std::cout << "Skipping, device " << device << " does not support cooperative groups"
<< std::endl;
return 0;
}
#endif
2. Initialize the cooperative group configuration
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
In the example, there is only one block in the grid, and the ``threads_per_block`` must be dividable with ``partition_size``.
.. code-block:: cpp
// Number of blocks to launch.
constexpr unsigned int num_blocks = 1;
// Number of threads in each kernel block.
constexpr unsigned int threads_per_block = 64;
// Total element count of the input vector.
constexpr unsigned int size = num_blocks * threads_per_block;
// Total elements count of a tiled_partition.
constexpr unsigned int partition_size = 16;
// Total size (in bytes) of the input vector.
constexpr size_t size_bytes = sizeof(unsigned int) * size;
static_assert(threads_per_block % partition_size == 0,
"threads_per_block must be a multiple of partition_size");
4. Launch the kernel
~~~~~~~~~~~~~~~~~~~~
The kernel launch is done with the ``hipLaunchCooperativeKernel`` of the cooperative groups API.
.. code-block:: cpp
void* params[] = {&d_vector, &d_block_reduced, &d_partition_reduced};
// Launching kernel from host.
HIP_CHECK(hipLaunchCooperativeKernel(vector_reduce_kernel<partition_size>,
dim3(num_blocks),
dim3(threads_per_block),
params,
0,
hipStreamDefault));\
// Check if the kernel launch was successful.
HIP_CHECK(hipGetLastError());
Conclusion
==========
With cooperative groups, you can easily use custom partitions to create custom tiles for custom solutions. You can find the complete code at `cooperative groups ROCm example. <https://github.com/ROCm/rocm-examples/tree/develop/HIP-Basic/cooperative_groups>`_
projects/hip/docs/tutorial/reduction.rst
+722
Melihat File
@@ -0,0 +1,722 @@
.. meta::
:description: HIP reduction tutorial
:keywords: AMD, ROCm, HIP, reduction, tutorial
*************************************************************
Reduction
*************************************************************
Reduction is a common algorithmic operation used in parallel programming to reduce an array of elements into a shorter array of elements or a single value. This document exploits reduction to introduce some key considerations while designing and optimizing GPU algorithms.
This document is a rejuvenation and extension of the invaluable `work of Mark Harris <https://developer.download.nvidia.com/assets/cuda/files/reduction.pdf>`_. While the author approaches the topic with a less naive approach, reviewing some original material is valuable to see how much the underlying hardware has changed. This document provides a greater insight to demonstrate progress.
The algorithm
=============
Reduction has many names depending on the domain; in functional programming it's referred to as `fold <https://en.wikipedia.org/wiki/Fold_(higher-order_function)>`_, in C++, it's called ``std::accumulate`` and in C++17, as ``std::reduce``. A reduction takes a range of inputs and "reduces" the given range with a binary operation to a singular or scalar output. Canonically, a reduction requires a "zero" element that bootstraps the algorithm and serves as one of the initial operands to the binary operation. The "zero" element is generally called `identity or neutral <https://en.wikipedia.org/wiki/Identity_element>`_ element in the group theory, which implies that it is an operand that doesn't change the result. Some typical use cases are: calculating a sum or normalizing a dataset and finding the maximum value in the dataset. The latter use case is discussed further in this tutorial.
.. figure:: ../data/tutorial/reduction/foldl.svg
:alt: Diagram demonstrating fold left
There are multiple variations of reduction that allow parallel processing. The approach taken by ``std::reduce`` requires the user-provided binary operator to operate on any combination of identity and input range elements, or even exclusively on any of them. This allows you to insert any number of identities to facilitate parallel processing and then combine the partial results of parallel execution.
.. figure:: ../data/tutorial/reduction/parallel_foldl.svg
:alt: Diagram demonstrating parallel fold left
Reduction on GPUs
=================
Implementing reductions on GPUs requires a basic understanding of the :doc:`/understand/programming_model`. The document explores aspects of low-level optimization best discussed through the :ref:`inherent_thread_model`, and refrains from using cooperative groups.
Synchronizing parallel threads of execution across a GPU is crucial for correctness as the partial results can't be synchronized before they manifest. Synchronizing all the threads running on a GPU at any given time is possible, however, it is a costly and intricate operation. If synchronization is not absolutely necessary, map the parallel algorithm so that multiprocessors and blocks can make independent progress and need not sync frequently.
There are ten reduction implementations in the `rocm-examples <https://github.com/ROCm/rocm-examples/tree/develop/Tutorials/reduction/include/Reduction>`_, which are described in the following sections.
Naive shared reduction
----------------------
The naive algorithm takes a tree-like shape, where the computational domain is purposefully distributed among blocks. In all blocks, all threads participate in loading data from persistent (from the kernel's perspective) global memory into the shared memory. This helps to perform tree-like reduction for a single thread by writing the partial result to global, in a location unique to the block, which allows the block to make independent progress. The partial results are combined in subsequent launches of the same kernel until a scalar result is reached.
.. figure:: ../data/tutorial/reduction/naive_reduction.svg
:alt: Diagram demonstrating naive reduction
This approach requires temporary storage based on the number of blocks launched, as each block outputs a scalar partial result. Depending on the need to store or destroy the input, a second temporary storage might be needed, which could be large enough to store the results of the second kernel launch. Alternatively, you can reuse the storage of the larger than necessary original input. These implementations differ so slightly that the document only considers the use case where the input could be destroyed.
.. code-block:: C++
std::size_t factor = block_size; // block_size from hipGetDeviceProperties()
auto new_size = [factor](const std::size_t actual)
{
// Every pass reduces input length by 'factor'. If actual size is not divisible by factor,
// an extra output element is produced using some number of zero_elem inputs.
return actual / factor + (actual % factor == 0 ? 0 : 1);
};
For threads that don't have unique inputs, feed ``zero_elem`` instances to threads. The backing of double-buffering is allocated as such:
.. code-block:: C++
// Initialize host-side storage
std::vector<unsigned> input(input_count);
std::iota(input.begin(), input.end(), 0);
// Initialize device-side storage
unsigned *front,
*back;
hipMalloc((void**)&front, sizeof(unsigned) * input_count);
hipMalloc((void**)&back, sizeof(unsigned) * new_size(input_count));
hipMemcpy(front, input.data(), input.size() * sizeof(unsigned), hipMemcpyHostToDevice);
Data is initialized on the host and dispatched to the device followed by the commencement of device-side reduction. The swapping of the double-buffer on the last iteration is omitted, therefore the result is in the back-buffer irrespective of the input size.
.. code-block:: C++
for (uint32_t curr = input_count; curr > 1;)
{
hipLaunchKernelGGL(
kernel,
dim3(new_size(curr)),
dim3(block_size),
factor * sizeof(unsigned),
hipStreamDefault,
front,
back,
kernel_op,
zero_elem,
curr);
curr = new_size(curr);
if (curr > 1)
std::swap(front, back);
}
This structure persists in the kernel throughout all the variations of reduction with slight modifications to ``factor`` and shared memory allocation:
.. code-block:: C++
template<typename T, typename F>
__global__ void kernel(
T* front,
T* back,
F op,
T zero_elem,
uint32_t front_size)
{
extern __shared__ T shared[];
// Overindex-safe read of input
auto read_global_safe = [&](const uint32_t i)
{
return i < front_size ? front[i] : zero_elem;
};
const uint32_t tid = threadIdx.x,
bid = blockIdx.x,
gid = bid * blockDim.x + tid;
// Read input from front buffer to shared
shared[tid] = read_global_safe(gid);
__syncthreads();
// Shared reduction
for (uint32_t i = 1; i < blockDim.x; i *= 2)
{
if (tid % (2 * i) == 0)
shared[tid] = op(shared[tid], shared[tid + i]);
__syncthreads();
}
// Write result from shared to back buffer
if (tid == 0)
back[bid] = shared[0];
}
While the ``tid % (2 * i) == 0`` indexing scheme yields correct results, it also leads to high thread divergence. Thread divergence indicates the event when the threads in a warp diverge, which implies that the threads have to execute different instructions in a given clock cycle. This is easily manifested using ``if-else`` statements as shown here, but can also be manifested as ``for`` loop dependent on thread ID lengths. Even though the number of active threads participating in the reduction reduces, warps remain active longer than necessary, as at least one lane in a warp hits the ``if`` statement.
Reducing thread divergence
--------------------------
You can reduce divergence by keeping dataflow between memory addresses identical but reassigning the thread ids.
.. figure:: ../data/tutorial/reduction/reduced_divergence_reduction.svg
:alt: Diagram demonstrating reduced divergence reduction
.. code-block:: diff
:emphasize-lines: 4-7
// Shared reduction
for (uint32_t i = 1; i < blockDim.x; i *= 2)
{
- if (tid % (2 * i) == 0)
- shared[tid] = op(shared[tid], shared[tid + i]);
+ if (uint32_t j = 2 * i * tid; j < blockDim.x)
+ shared[j] = op(shared[j], shared[j + i]);
__syncthreads();
}
This way inactive threads start accumulating uniformly towards the higher thread ID index range and might uniformly skip to ``__syncthreads()``. However, this introduces a bank conflicts issue.
Resolving bank conflicts
------------------------
Both AMD and NVIDIA implement shared memory in the hardware by organizing storage into banks of various sizes. This hardware element is known as Local Data Share (LDS) on AMD hardware. On NVIDIA hardware, it's implemented using the same silicon as the L1 data cache. You can think of shared memory as a striped 2-dimensional range of memory. Shared memory bank's count, width, and depth depend on the architecture. A bank conflict occurs when different threads in a warp access the same bank during the same operation. In this case, the hardware prevents the attempted concurrent accesses to the same bank by converting them into serial accesses.
- `"AMD Instinct MI200" Instruction Set Architecture, Chapter 11.1 <https://www.amd.com/content/dam/amd/en/documents/instinct-tech-docs/instruction-set-architectures/instinct-mi200-cdna2-instruction-set-architecture.pdf>`_
- `"RDNA 2" Instruction Set Architecture, Chapter 10.1 <https://www.amd.com/content/dam/amd/en/documents/radeon-tech-docs/instruction-set-architectures/rdna2-shader-instruction-set-architecture.pdf>`_
A notable exception is when the shared read uniformly broadcasts to the same address across the entire warp. A better implementation of the naive algorithm is to form continuous ranges of the threads activities and their memory accesses.
.. code-block:: diff
:emphasize-lines: 2-7
// Shared reduction
-for (uint32_t i = 1; i < blockDim.x; i *= 2)
-{
- if (tid % (2 * i) == 0)
+for (uint32_t i = blockDim.x / 2; i != 0; i /= 2)
+{
+ if (tid < i)
shared[tid] = op(shared[tid], shared[tid + i]);
__syncthreads();
}
.. figure:: ../data/tutorial/reduction/conflict_free_reduction.svg
:alt: Diagram demonstrating bank conflict free reduction
.. note::
To avoid bank conflicts, read shared memory in a coalesced manner, which implies that reads/writes of each lane in a warp evaluate to consecutive locations. Analyzing the read/write patterns could help you to understand the cause of bank conflicts. For more details, check `CDNA3 ISA <https://www.amd.com/content/dam/amd/en/documents/instinct-tech-docs/instruction-set-architectures/amd-instinct-mi300-cdna3-instruction-set-architecture.pdf>`_ or `RDNA3 ISA <https://www.amd.com/content/dam/amd/en/documents/radeon-tech-docs/instruction-set-architectures/rdna3-shader-instruction-set-architecture-feb-2023_0.pdf>`_ data share operations chapter.
Utilize upper half of the block
-------------------------------
The preceding implementation is free of low-level GPU-specific anti-patterns. However, it still exhibits some common shortcomings. The loop performing the reduction in the shared memory starts from ``i = blockDim.x / 2`` and the first predicate ``if (tid < i)`` immediately disables half of the block, which only helps load the data into the shared memory. You can change the kernel along with the calculation of ``factor`` on the host, as shown here:
.. code-block:: diff
:emphasize-lines: 3,4
const uint32_t tid = threadIdx.x,
bid = blockIdx.x,
- gid = bid * blockDim.x + tid;
+ gid = bid * (blockDim.x * 2) + tid;
// Read input from front buffer to shared
-shared[tid] = read_global_safe(gid);
+shared[tid] = op(read_global_safe(gid), read_global_safe(gid + blockDim.x));
__syncthreads();
By eliminating half of the threads and giving meaningful work to all the threads by unconditionally performing a binary ``op``, you can prevent the wastage of half of the threads.
Even though global memory is read in a coalesced fashion, as preferred by the memory controller, optimal performance is still limited by the instruction throughput.
Omit superfluous synchronization
--------------------------------
Warps are known to execute in a strict lockstep fashion. Therefore, once shared reduction reaches a point where only a single warp participates meaningfully, you can cut short the loop and let the rest of the warps terminate. Moreover, you can also unroll the loop without syncing the entire block.
The ``tmp`` namespace used beyond this point in this document holds a handful of template meta-programmed utilities to facilitate writing flexible and optimal code.
:code:`tmp::static_for` is not just a constant folding within the optimizer but a variation of the language :code:`for` loop, where the running index is a compile-time constant and is eligible for use in compile-time evaluated contexts.
Consider the following code:
.. code-block:: C++
constexpr int size = 4;
for (int i = 0 ; i < size ; ++i)
{
printf("%d", i);
}
This compiles to the following binaries:
**LLVM Block**
.. code-block::
main:
push rbx
lea rbx, [rip + .L.str]
mov rdi, rbx
xor esi, esi
xor eax, eax
call printf@PLT
mov rdi, rbx
mov esi, 1
xor eax, eax
call printf@PLT
mov rdi, rbx
mov esi, 2
xor eax, eax
call printf@PLT
mov rdi, rbx
mov esi, 3
xor eax, eax
call printf@PLT
xor eax, eax
pop rbx
ret
.L.str:
.asciz "%d"
**GCC**
.. code-block:: asm
.LC0:
.string "%d"
main:
push rbx
xor ebx, ebx
.L2:
mov esi, ebx
mov edi, OFFSET FLAT:.LC0
xor eax, eax
add ebx, 1
call printf
cmp ebx, 4
jne .L2
xor eax, eax
pop rbx
ret
**MSVC**
.. code-block::
main PROC
$LN12:
push rbx
sub rsp, 32
xor ebx, ebx
npad 8
$LL4@main:
mov edx, ebx
lea rcx, OFFSET FLAT:'string'
call printf
inc ebx
cmp ebx, 4
jl SHORT $LL4@main
xor eax, eax
add rsp, 32
pop rbx
ret 0
main ENDP
LLVM unrolls the loop and compiles to a flat series of ``printf`` invocations, while both GCC and MSVC keep the loop intact, as visible from the compare (``cmp``) and the jump (``jne``, ``jl``) instructions. LLVM code generation is identical to manually writing the unrolled loop:
.. code-block:: C++
printf("%d", 0);
printf("%d", 1);
printf("%d", 2);
printf("%d", 3);
While various non-standard pragmas are available to hint or force the compiler to unroll the loop, we instead use template meta-programming to force feed the compiler the unrolled loop.
.. code-block:: C++
constexpr int size = 4;
// Maybe unrolled loop
for (int i = 0 ; i < size ; ++i)
{
printf("%d", i);
}
// Force unrolled loop
using namespace tmp;
static_for<0, less_than<size>, increment<1>>([]<int i>()
{
printf("%d", i);
});
The most notable structural difference is that in the language ``for`` loop, the loop variable is given a name in the beginning, while in the ``static_for`` utility, the loop variable is given a name in the end. An important bonus is that in the loop's body, you can use the running index ``i`` in contexts requiring constant expressions such as template arguments or inside ``if constexpr``.
:code:`tmp::static_switch` takes runtime value and runtime dispatches to a range of set of tabulated functions, where said value is a compile-time constant and is eligible for use in compile-time evaluated contexts.
Consider the following code:
.. code-block:: C++
int warp_size = device_props.warpSize;
switch (warp_size)
{
case 32:
hipLaunchKernelGGL(kernel<32>, ...);
break;
case 64:
hipLaunchKernelGGL(kernel<64>, ...);
break;
}
In the preceding code, note the code repetition for all possible values of ``warp_size``, the code is prepared to handle. To avoid this, use ``tmp::static_switch``, as shown:
.. code-block:: C++
tmp::static_switch<std::array{32, 64}>(warp_size, [&]<int WarpSize>()
{
hipLaunchKernelGGL(kernel<WarpSize>, ...);
});
.. code-block:: diff
:emphasize-lines: 1,2,9,10,16-24
-template<typename T, typename F>
+template<uint32_t WarpSize, typename T, typename F>
__global__ void kernel(
...
)
{
...
// Shared reduction
-for (uint32_t i = blockDim.x / 2; i != 0; i /= 2)
+for (uint32_t i = blockDim.x / 2; i > WarpSize; i /= 2)
{
if (tid < i)
shared[tid] = op(shared[tid], shared[tid + i]);
__syncthreads();
}
+// Warp reduction
+tmp::static_for<WarpSize, tmp::not_equal<0>, tmp::divide<2>>([&]<int I>()
+{
+ if (tid < I)
+ shared[tid] = op(shared[tid], shared[tid + I]);
+#ifdef __HIP_PLATFORM_NVIDIA__
+ __syncwarp(0xffffffff >> (WarpSize - I));
+#endif
+});
Because HIP typically targets hardware with warp sizes of 32 (NVIDIA GPUs and RDNA AMD GPUs) and 64 (CDNA AMD GPUs), portable HIP code must handle both. That is why instead of assuming a warp size of 32, make the warp size a template argument of the kernel. This allows you to unroll the final loop using ``tmp::static_for`` in a parametric way but still having the code read much like an ordinary loop.
Promoting the warp size to being a compile-time constant also requires you to handle it similarly on the host-side. You can sandwich the kernel launch with ``tmp::static_switch``, promoting the snake-case run-time ``warp_size`` variable to a camel-case compile-time constant ``WarpSize``.
.. code-block:: diff
:emphasize-lines: 4,5,7,8,18
// Device-side reduction
for (uint32_t curr = input_count; curr > 1;)
{
+ tmp::static_range_switch<std::array{32, 64}>(warp_size, [&]<int WarpSize>() noexcept
+ {
hipLaunchKernelGGL(
- kernel,
+ kernel<WarpSize>,
dim3(new_size(curr)),
dim3(block_size),
factor * sizeof(unsigned),
hipStreamDefault,
front,
back,
kernel_op,
zero_elem,
curr);
+ });
...
}
.. note::
Neither RDNA- nor CDNA-based AMD hardware provides guaranteed independent progress to lanes of the same warp. When targeting NVIDIA hardware, lanes of a warp might execute somewhat independently as long as the programmer assists the compiler using dedicated built-in functions. This feature is called Independent Thread Scheduling. The HIP headers don't expose the necessary warp primitives and their overloads.
Portable applications can still tap into this feature with carefully ``#ifdef`` -ed code, but at this particular optimization level, it's a requirement. The code implicitly relies on the lockstep behavior of an ROCm wavefront, but CUDA warps don't share this property. You must synchronize all the active lanes of a warp to avoid a data race with some lanes progressing faster than others in the same warp.
Unroll all loops
----------------
While the previous step primarily aims to remove unnecessary syncing, it also unrolls the end of the loop. However, you could also force unrolling the first part of the loop. This saves a few scalar registers (values the compiler can prove to be uniform across warps).
.. code-block:: diff
:emphasize-lines: 1-4,11,12,17,18,20-23,26
-template<uint32_t WarpSize, typename T, typename F>
-__global__ void kernel(
+template<uint32_t BlockSize, uint32_t WarpSize, typename T, typename F>
+__global__ __launch_bounds__(BlockSize) void kernel(
T* front,
T* back,
F op,
T zero_elem,
uint32_t front_size)
{
- extern __shared__ T shared[];
+ __shared__ T shared[BlockSize];
...
// Shared reduction
- for (uint32_t i = blockDim.x / 2; i > WarpSize; i /= 2)
+ tmp::static_for<BlockSize / 2, tmp::greater_than<WarpSize>, tmp::divide<2>>([&]<int I>()
{
- if (tid < i)
- shared[tid] = op(shared[tid], shared[tid + i]);
+ if (tid < I)
+ shared[tid] = op(shared[tid], shared[tid + I]);
__syncthreads();
}
+ );
Introducing yet another template argument for the kernel and moving from ``for`` to ``tmp::static_for`` leads to the following two notable improvements:
- Introducing new attribute ``__launch_bounds__(BlockSize)`` to the kernel instructs the compiler that the kernel will only be launched using the designated block size. This implies that the launches of differing block sizes will fail. This allows the optimizer to enroll the ``blockDim.x`` variable in constant folding as well as get information about register usage.
- Turning the block size into a compile-time constant allows you to statically allocate the shared memory.
Communicate using warp-collective functions
-------------------------------------------
Shared memory provides a fast communication path within a block, however when performing reduction within the last warp, you can use faster means of communication, which is warp-collective or cross-lane functions. Instead of using the hardware-backed shared memory, you can directly copy between the local memory (registers) of each lane in a warp. This can be achieve using the shuffle functions.
See how to use ``__shfl_down()``, which is one of the most restrictive but also the most structured communication schemes.
.. code-block:: C++
// Warp reduction
if (tid < WarpSize)
{
T res = op(shared[tid], shared[tid + WarpSize]);
tmp::static_for<WarpSize / 2, tmp::not_equal<0>, tmp::divide<2>>([&]<int Delta>()
{
res = op(res, __shfl_down(res, Delta));
});
// Write result from shared to back buffer
if (tid == 0)
back[bid] = res;
}
Using warp-collective functions for communication requires the control flow to be uniform across warps, as the name warp-collective implies. Therefore, you can see that the thread ID is being checked outside the loop, but the result is written inside due to variable scoping.
Prefer warp communication over shared
-------------------------------------
As mentioned in the previous step, communication between local memory is faster than shared memory. Instead of relying on the local memory only at the end of the tree-like reduction, a better approach is to turn the tree reduction inside out and perform multiple warp reductions in parallel on all active threads, thus communicating only their partial results through the shared memory.
.. figure:: ../data/tutorial/reduction/warp_reduction.svg
:alt: Diagram demonstrating warp reduction
.. figure:: ../data/tutorial/reduction/warp_reduction_with_shared.svg
:alt: Diagram demonstrating warp reduction and results store in shared memory
The kernel versions differ significantly enough to be described using a diff; use afresh instead.
.. code-block:: C++
template<uint32_t BlockSize, uint32_t WarpSize, typename T, typename F>
__global__ __launch_bounds__(BlockSize) void kernel(
T* front,
T* back,
F op,
T zero_elem,
uint32_t front_size)
{
// ...
}
The kernel signature and the reduction factor are the same as in previous cases; only the implementation differs.
.. code-block:: C++
static constexpr uint32_t WarpCount = BlockSize / WarpSize;
__shared__ T shared[WarpCount];
auto read_global_safe =
[&](const uint32_t i) { return i < front_size ? front[i] : zero_elem; };
auto read_shared_safe =
[&](const uint32_t i) { return i < WarpCount ? shared[i] : zero_elem; };
const uint32_t tid = threadIdx.x,
bid = blockIdx.x,
gid = bid * (blockDim.x * 2) + tid,
wid = tid / WarpSize,
lid = tid % WarpSize;
// Read input from front buffer to local
T res = op(read_global_safe(gid), read_global_safe(gid + blockDim.x));
As we communicate the results of warps through shared memory, the same number of elements are required in the shared memory as warps within the block. Similar to how you can only launch kernels at block granularity, you can only warp reduce with ``WarpSize`` granularity due to the collective nature of the cross-lane builtins. To address this, you can use ``read_shared_safe`` to pad overindexing by reading ``zero_elem``. Reading from global remains unaffected.
.. code-block:: C++
// Perform warp reductions and communicate results via shared
// for (uint32_t ActiveWarps = WarpCount;
// ActiveWarps != 0;
// ActiveWarps = ActiveWarps != 1 ?
// divide_ceil(ActiveWarps, WarpSize) :
// ActiveWarps = 0)
tmp::static_for<
WarpCount,
tmp::not_equal<0>,
tmp::select<
tmp::not_equal<1>,
tmp::divide_ceil<WarpSize>,
tmp::constant<0>>>([&]<uint32_t ActiveWarps>()
{
if(wid < ActiveWarps)
{
// Warp reduction
tmp::static_for<WarpSize / 2, tmp::not_equal<0>, tmp::divide<2>>([&]<int Delta>()
{
res = op(res, __shfl_down(res, Delta));
});
// Write warp result from local to shared
if(lid == 0)
shared[wid] = res;
}
__syncthreads();
// Read warp result from shared to local
res = read_shared_safe(tid);
});
// Write result from local to back buffer
if(tid == 0)
back[bid] = res;
``ActiveWarps`` iterates from ``WarpCount`` until it reaches ``0``. Every iteration of ``ActiveWarps`` reduces the ``WarpSize``. In cases where the partial result count isn't a divisor of ``ActiveWarps`` and you need to launch an extra warp, use ``tmp::divide_ceil``, which always rounds to positive infinity. The tertiary ``tmp::select`` is required because such division never reaches ``0``, so you must terminate the loop after the last warp concludes.
In each iteration, if the warp is active, which means it has at least a single valid input, it carries out a pass of warp reduction and writes output based on warp ID. Reading is carried out based on thread ID. Global output continues to be based on block ID.
Amortize bookkeeping variable overhead
--------------------------------------
The previous sections explained how to reduce register usage to improve occupancy. This allows more blocks to execute in parallel on all multiprocessors, leading to more global store/load latency to be hidden. Reducing the number of kernels in flight while still carrying out the same workload reduces the wastage of registers while loading and maintaining bookkeeping variables such as kernel indices.
An example of this optimization is performing one binary ``op`` while loading input from global. Even though the operation is said to be carried out "in flight", the two values are loaded into local memory (registers) before ``op`` is called.
A more general form of this optimization is wrapping most kernel logic in loops that carry out the workload of multiple kernel instances but require storing only a single instance of most of the bookkeeping logic. In code, this multiplicity factor is referred to via the ``ItemsPerThread`` compile-time constant, which is supplied by a template argument to allow for loop unrolling.
This kernel variant utilizes another generally applicable utility known as ``hip::static_array``, which is a more restrictive wrapper over the builtin array than ``std::array``, as it allows indexing only compile-time constants using the usual tuple-like ``template <size_t I> auto get<I>(...)`` interface.
.. note::
On a GPU, there is no stack, and the local memory is provisioned from the register file. This provisioning takes place statically. To paraphrase, the address range of a thread's local memory is determined at compile-time. When an array is defined and used in the local storage, the compiler can only maintain its storage in the register file as long as all accesses to the array are computable by the compiler at compile-time. It doesn't need to be a compile-time constant as long as the compiler can resolve the addresses of the accesses through constant folding or some other means. If the compiler fails to do so, the array will be backed by global memory, which is indicated by allocating a non-zero number of spill registers observable using static analysis tools. However, this is slower by the magnitude of multiple order. ``hip::static_array`` via its ``hip::get<>`` interface ensures that no such spills occur.
.. code-block:: C++
template<uint32_t BlockSize, uint32_t WarpSize, uint32_t ItemsPerThread>
__global__ static __launch_bounds__(BlockSize) void kernel(...)
The kernel now has three compile-time configurable parameters. The only part of the kernel that changes depends on how you load data from global and perform the binary operation on those loaded values. So, the following step to read input from front buffer to global is now split into two steps: :ref:`reading-items` and :ref:`processing-items` .
.. code-block:: C++
// Read input from front buffer to local
T res = op(read_global_safe(gid), read_global_safe(gid + blockDim.x));
.. _reading-items:
Reading ``ItemsPerThread``
^^^^^^^^^^^^^^^^^^^^^^^^^^
The change to reading happens inside `read_global_safe`:
.. code-block:: C++
auto read_global_safe = [&](const int32_t i) -> hip::static_array<T, ItemsPerThread>
{
return [&]<int32_t... I>(std::integer_sequence<int32_t, I...>)
{
if(i + ItemsPerThread < front_size)
return hip::static_array<T, ItemsPerThread>{
front[i + I]...
};
else
return hip::static_array<T, ItemsPerThread>{
(i + I < front_size ? front[i + I] : zero_elem)...
};
}(std::make_integer_sequence<int32_t, ItemsPerThread>());
};
Note that each array element is being loaded consecutively without the flexibility of a configurable ``ItemsPerThread`` property. This is morally equivalent to:
.. code-block:: C++
T arr[4] = {
front[gid + 0],
front[gid + 1],
front[gid + 2],
front[gid + 3]
}
This is exactly what's happening in the ``front[i + I]...`` fold-expression. However, this can only be issued if the entire read operates on real input without padding using ``zero_elem``. If some reads over-index the input, the read turns into:
.. code-block:: C++
T arr[4] = {
i + 0 < front_size ? front[i + 0] : zero_elem,
i + 1 < front_size ? front[i + 1] : zero_elem,
i + 2 < front_size ? front[i + 2] : zero_elem,
i + 3 < front_size ? front[i + 3] : zero_elem
}
This makes it easier for the compiler to recognize vector loads from global. As the performance at large is dominated by how you move the data, it's only natural to utilize dedicated instructions to move more data with less binary. This is evident by the huge performance improvement when loading two values per thread. For more information, see `the compiler explorer <https://godbolt.org/z/b36Eea69q>`_ to learn how loading for AMD (both RDNA and CDNA) compiles to ``global_load_dwordx4``, where ``x4`` denotes the 4-vector variant of the instruction.
.. note::
Note that ``read_global_safe``, which used to take an ``uint32_t`` as the index type, now takes a signed integer. When indexing an array with unsigned integers, the compiler has to handle integer overflows, as the C/C++ standards defined them. It might happen that some part of the vector load indices overflow, thus resulting in a non-contiguous read. If you change the previously linked code to use an unsigned integer as the thread ID, the compiler won't emit a vector load. Signed integer overflow is an undefined behavior, and hence, unknown to the optimizer. To convey the absence of overflow to the compiler with unsigned indices, add ``__builtin_assume(gid + 4 > gid)``, or the more portable ``[[assume]](gid + 4 > gid)``, once ``amdclang++`` supports it.
``read_global_safe`` implementation is an Immediately Invoked Lambda Expression (IILE), because ``ItemsPerThread`` is an integer value, while you need a compile-time ``iota``-like sequence of integers as a pack for the fold-expressions to expand on. This can only occur as part of template argument deduction on the IILE.
.. _processing-items:
Processing ``ItemsPerThread``
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Once the kernel reads ``ItemsPerThread`` number of inputs to local, it immediately reduces them to a scalar. There is no reason to propagate the input element multiplicity to the warp reduction phase.
.. code-block:: C++
T res = [&]()
{
// Read input from front buffer to local
hip::static_array<T, ItemsPerThread> arr = read_global_safe(gid);
// Reduce ItemsPerThread to scalar
tmp::static_for<1, tmp::less_than<ItemsPerThread>, tmp::increment<1>>([&]<int I>()
{
get<0>(arr) = op(get<0>(arr), get<I>(arr));
});
return get<0>(arr);
}();
Two-pass reduction
------------------
Alter kernel launch and input fetching such that no more blocks are launched than what a subsequent kernel launch's single block can conveniently reduce, while performing multiple passes of input reading from global and combining their results before engaging in the end game tree-like reduction.
With this method, you can save at least one to two kernel launches for large inputs.
Global data share
-----------------
.. warning::
This modification can only be executed on AMD hardware.
Perform the first step of the two-pass reduction, but in the end, instead of writing to global and reading it back in a subsequent kernel, write the partial results to the Global Data Share (GDS). This is an ``N+1`` th shared memory that is accessed by all multiprocessors and is also on-chip memory.
.. note::
The API doesn't guarantee the order in which blocks are scheduled even though all GPUs schedule them in the same monotonically increasing order of block ids. Relying on this implicitly, the last block of a grid is in the optimal position to observe the side effects of all other blocks (using spinlocks or other methods) without occupying a multiprocessor for longer than necessary.
Without launching a second kernel, you can make the last block collect the results of all other blocks from GDS by implicitly exploiting the scheduling behavior or relying on another AMD-specific feature called Global Wave Sync (GWS) to merge them for a final tree-like reduction.
.. note::
GDS and GWS are reserved runtime features that the HIP API doesn’t cover. Invoking these functionalities requires inline AMDGCN assembly. Moreover, the fact that the runtime doesn’t virtualize the GDS, imposes further restrictions on concurrent scheduling of other kernels.
Conclusion
==========
Optimizing code on GPUs, like on any other architecture, requires careful consideration and balancing of resources and costs of various operations to obtain optimal performance. This document explored optimizing reductions much beyond the territory of diminishing returns. This approach introduced multiple optimization techniques and discussed opportunities.
The document focused on reductions when an entire device participates in it. Still, the choice of optimal compile-time constants or even the algorithm itself might not be optimal when its multiple blocks participate in multiple parallel reductions or when each thread performs its reduction. However, when multiple devices participate in the same reduction, other aspects must be considered.
Most solutions, including the ones covered in this document, are given to the end users in a turnkey fashion via algorithm primitive libraries. These solutions might not be the fastest in all cases, but they are close to being the gold standard for carrying out certain operations as reasonably as possible.

Some files were not shown because too many files have changed in this diff Show More