033d799524
2.7.8-1
...
Fix collective mismatch error when using ncclSend/ncclRecv
2020-07-27 16:34:09 -07:00
e7a10aa0e4
Topology tuning for 4P2H on Rome ( #242 )
...
* Topology tuning for 4P2H on Rome
* Use ncclTopoIdToIndex
2020-07-27 11:53:57 -07:00
8d5fb920b6
ib-test: support multiple channels ( #241 )
2020-07-27 11:03:12 -07:00
fe3d520601
Merge pull request #240 from ROCmSoftwarePlatform/sourav/topo-expl-1
...
simplify model definitions in topo expl
2020-07-22 12:35:17 -05:00
2475daafee
add 4 node 8P6L 1 NIC 2nd Hive model
2020-07-22 16:27:15 +00:00
db55afb014
simplify model definitions in topo expl
2020-07-22 16:05:53 +00:00
d5f90e19b5
Add 8P6L multi-node models ( #239 )
2020-07-21 14:10:36 -07:00
684f3e6af4
Adding better naming to unit tests for filtering; adding short and full unit test suites ( #235 )
2020-07-21 12:19:47 -06:00
35c5a7fe45
Fix RCCL build package name ( #236 )
2020-07-20 14:43:00 -07:00
99a491273f
Changing GTest inclusion in cmake to use find_package ( #234 )
...
* GTest is used via find_package. No longer downloaded in cmake.
* Adding error handling
2020-07-15 20:51:48 -06:00
7f93aa7e53
Changing dependency to hip-rocclr ( #228 )
2020-07-14 17:49:56 -06:00
ab787c767e
Change default channels duplication for chordal ring ( #233 )
2020-07-14 15:16:50 -07:00
f87ba17737
Removing UnitTest as install, removing unused env var ( #231 )
2020-07-10 09:30:28 -06:00
5215130168
Revert "Split primitive class to smaller structures" ( #230 )
...
This reverts commit 486fd436af
2020-07-08 11:06:50 -07:00
1addf4f196
Match RCCL package name to API version ( #229 )
2020-07-07 13:30:39 -07:00
2d8601701d
Fix build action order
...
Add $(INCTARGETS) to build dependencies of %.o and $(DEVICELIB).
As there were no dep files during the first build, Make may kick off source
compilation before nccl.h got generated, which leads to occasional build
failures on systems with high core count. The build failure could be
reproduced reliably with a `sleep 5` in $(INCDIR)/nccl.h rule.
2020-07-07 10:20:51 -07:00
9bd4c14603
Adding appropriate references in rccl-prim-test ( #227 )
...
Adding appropriate references to rccl-prim-test.
2020-07-06 10:15:03 -06:00
ecae1cd76a
Merge pull request #226 from wenkaidu/develop
...
Sync up to NCCL 2.7.6
2020-07-06 09:10:09 -07:00
da3b197d6c
Merge remote-tracking branch 'nccl/master' into develop
2020-07-01 16:51:25 -07:00
d3548cc474
topo_expl: each rank needs to have its own memory for graphs ( #225 )
2020-07-01 15:11:02 -07:00
a6be82f5ab
topo_expl: fix broken build ( #224 )
2020-06-30 11:11:23 -07:00
a144a85465
Merge pull request #223 from wenkaidu/sendrecv
...
Use separate threads for send and receive
2020-06-30 10:50:06 -07:00
8db0aa8f4c
gtest: extend testing up to 8 GPUs
2020-06-29 09:32:31 -07:00
964c4c2061
Merge sendrecv kernel from NCCL 2.7.3
...
This commit was cherry-picked and modified from
https://github.com/NVIDIA/nccl/commit/5949d96f36d050e59d05872f8bbffd2549318e95
2020-06-29 08:47:46 -07:00
b90735c935
Use separate threads for send and receive
2020-06-29 08:47:15 -07:00
1952325569
2.7.6-1
...
Fix crash when NVswitch is not visible inside a VM.
2020-06-26 16:35:54 -07:00
01afd20a77
2.7.5-1
...
Minor fixes for A100 platforms.
Add a WARN for invalid GroupEnd call.
2020-06-26 14:39:49 -07:00
84f8ba3bb0
Revert use posix_memalign for network buffer allocation on host memory ( #222 )
2020-06-24 11:25:55 -07:00
0eb19a563a
Use posix_memalign for network buffer allocation on host memory ( #221 )
...
* Use posix_memalign for network buffer allocation on host memory
* ib-test: add ability to specify run iterations
* ib-test: define iterations as multiple of default cycles
* Add checking to posix_memalign return value
2020-06-22 13:06:25 -07:00
dc739c4e70
ib-test: support host memory allocation through posix_memalign ( #220 )
...
* ib-test: support host memory allocation through posix_memalign
* ib-test: add missing CUDACHECK to hip calls
2020-06-17 16:16:54 -07:00
8d21adb5e3
Documentation updates for NCCL 2.7.0 ( #219 )
...
* Making hip-clang the default compiler; documentation update
* Adding back --hip-clang to install.sh as a silent option for CI
* Documentation updates for NCCL 2.7
* Restoring deleted line in install script
2020-06-16 16:48:11 -06:00
0023b9b081
Merge pull request #218 from wenkaidu/develop
...
Develop
2020-06-16 13:30:50 -07:00
cfa97eccd3
Add IB/RDMA unit test
2020-06-16 18:29:17 +00:00
95b8f70d15
Limit network profiling support to simple protocol and avoid overflow
2020-06-15 20:51:36 +00:00
7484e53ff7
Rework network proxy profiling
2020-06-13 03:13:58 +00:00
b257676f30
Reduce RCCL kernel count as we don't pass first coll in argument
2020-06-12 21:30:04 +00:00
a6d621176c
Sender rank's opCount maybe ahead by one if it finishes earlier
2020-06-12 03:39:45 +00:00
fee1a20b74
gtest: add scatter, gather and all to all unit tests
2020-06-09 17:44:15 -07:00
e98891d039
Log NUMA node of RDMA host buffer allocation
2020-06-09 17:44:15 -07:00
812543104d
Add network proxy profiling support
2020-06-09 17:44:15 -07:00
c9aa11928a
Calculate and use total wait cycles for RCCL profiling
2020-06-09 17:44:15 -07:00
e80e29573c
Add gather, scatter and alltoall collectives
...
Introducing 3 new APIs:
ncclResult_t ncclGather(const void* sendbuff, void* recvbuff, size_t sendcount,
ncclDataType_t datatype, int root, ncclComm_t comm, hipStream_t stream);
ncclResult_t ncclScatter(const void* sendbuff, void* recvbuff,
size_t recvcount, ncclDataType_t datatype, int root, ncclComm_t comm,
hipStream_t stream);
ncclResult_t ncclAllToAll(const void* sendbuff, void* recvbuff, size_t count,
ncclDataType_t datatype, ncclComm_t comm, hipStream_t stream);
Only out of place operation is supported.
Preprocessor symbol RCCL_GATHER_SCATTER=1 indicates API availibility.
By default the APIs launche RCCL kernel implementation, which can be disabled by
RCCL_ALLTOALL_KERNEL_DISABLE=1. Then the APIs use wrapper around ncclSend and ncclRecv.
2020-06-09 17:44:08 -07:00
71ec3e09df
tpol_expl: update to 2.7
2020-06-09 17:40:24 -07:00
26a0fd2517
Merge remote-tracking branch 'nccl/master' into develop
2020-06-09 17:40:11 -07:00
5949d96f36
2.7.3-1
...
Add support for A100 GPU and related platforms.
Add support for CUDA 11.
Add support for send/receive operations (beta).
2020-06-08 09:31:44 -07:00
87db65f22d
Fixing CI as install.sh script should not install dependencies without user request ( #217 )
2020-06-05 11:04:03 -06:00
dc403e0ca2
Making hip-clang the default compiler; documentation update ( #216 )
...
* Making hip-clang the default compiler; documentation update
* Adding back --hip-clang to install.sh as a silent option for CI
2020-06-04 11:58:27 -06:00
2a4514772c
Merge pull request #214 from wenkaidu/gdr
...
Use cached value for detecting GDR support only once
2020-05-22 13:36:23 -07:00
67c8e72ce3
Use cached value for detecting GDR support only once
2020-05-22 17:19:10 +00:00
957be85944
Merge pull request #212 from wenkaidu/version
...
Report HIP version in logs
2020-05-20 16:25:54 -07:00
David Addison
Wenkai Du
Sourav Chakraborty
Stanley Tsang
saadrahim
gilbertlee-amd
Riatre Foo
Wenkai Du
Sylvain Jeaugey