Wenkai Du
d2fbcfea02
Misc fixes and improvements for 2.5.6
...
1. Fix RCCL unit test
2. Add ROME detection and tuning
3. Change default P2P level
4. Fix search algorithm for XGMI
5. Remove explicit channel duplication with implicit by using half of link speed
6. Add collective trace support
7. Correct Intel Skylake CPU detection and bandwidth
8. Fix topo connect function
9. Disable GDR read and remove unreachable code
10. Disable LL128 kernels
11. Add tuning parameters
12. Use original clock64() implementation which returns RTC counter value
13. Print out timestamp of collective trace
14. Do not use struct ncclColl in kernel launch parameter
15. Fix abort handling and add tracing
17. Add __launch_bounds__ to kernel functions
18. Remove unused abortCount
19. Unset default MIN_NRINGS and MIN_NCHANNELS
20. Do not allocate shared memory when not using LL128 kernels
21. Correct time print out in tuning log
[ROCm/rccl commit: 1e55645d97 ]
2020-01-29 15:27:05 -08:00
Wenkai Du
d7d4175df0
Merge remote-tracking branch 'remotes/nccl/master' into rccl_2.5.6
...
[ROCm/rccl commit: 6648c81dc6 ]
2019-12-03 15:42:04 -08:00
Wenkai Du
4cb52294d5
Disable direct buffers to reduce scratch memory size
...
[ROCm/rccl commit: a0be2b8812 ]
2019-11-20 13:03:16 -08:00
Sylvain Jeaugey
71560fd67b
2.5.6-1 ( #255 )
...
Add LL128 Protocol.
Rewrite the topology detection and tree/ring creation (#179 ). Improve
tree performance by sending/receiving from different GPUs. Add
model-based tuning to switch between the different algorithms and
protocols.
Rework P2P/SHM detection in containers (#155 , #248 ).
Detect duplicated devices and return an error (#231 ).
Add tuning for GCP
[ROCm/rccl commit: 299c554dcc ]
2019-11-19 14:57:39 -08:00
Wenkai Du
55ad3c801e
Support bfloat16 on rest of the unit tests
...
[ROCm/rccl commit: 4ca05c1297 ]
2019-11-18 14:18:34 -08:00
Wenkai Du
7dc39b8928
Add bfloat16 all reduce unit test
...
[ROCm/rccl commit: bdac0256a5 ]
2019-11-18 13:50:29 -08:00
Wenkai Du
1e182391ad
Add bfloat16 support in RCCL
...
Preprocessor symbol RCCL_BFLOAT16 is used as feature indicator
[ROCm/rccl commit: 5e109ed400 ]
2019-11-18 13:45:53 -08:00
Wenkai Du
8b1ce44c2a
Temporary disable 0x803 target due to build error
...
[ROCm/rccl commit: cd7ab1425b ]
2019-11-14 11:17:41 -08:00
Wenkai Du
8e60479385
Merge pull request #151 from wenkaidu/prim_test
...
rccl_prim_test: Generalize ring topology and duplications
[ROCm/rccl commit: 55c07e4fb7 ]
2019-11-13 08:17:55 -08:00
Siu Chi Chan
9b19999918
Bump up HCC version for -hc-function-calls switch
...
[ROCm/rccl commit: 08ba92f1b0 ]
2019-11-12 14:16:35 -05:00
Wenkai Du
25b3175e82
rccl_prim_test: Generalize ring topology and duplications
...
Allow user specified ring topology from command line and duplicated
to requested number of workgroups:
./rccl_prim_test -w 12 -p copy -r "0 1 2 3|3 2 1 0|0 2 1 3|3 1 2 0|0 2 3 1|1 3 2 0"
[ROCm/rccl commit: 07bb6fce8f ]
2019-11-11 15:42:24 -08:00
Wenkai Du
66e4337c6e
Merge pull request #149 from wenkaidu/rtc
...
Correct RTC frequencies for profiling purpose
[ROCm/rccl commit: 277c72a638 ]
2019-11-06 08:02:58 -08:00
gilbertlee-amd
22cbbb9004
Adding interactive mode for profiling purposes ( #150 )
...
[ROCm/rccl commit: fd94f4fa25 ]
2019-11-05 17:10:16 -07:00
Wenkai Du
62042e47bc
Correct RTC frequencies for profiling purpose
...
[ROCm/rccl commit: 8995047830 ]
2019-11-05 11:36:45 -08:00
Wenkai Du
41f6319b33
Check for fine grain support using memory allocation
...
[ROCm/rccl commit: 669f1951a4 ]
2019-11-01 15:58:49 -07:00
Wenkai Du
0d6b476b08
Merge pull request #145 from wenkaidu/prim_test
...
rccl-prim-test: use hipExtLaunchMultiKernelMultiDevice and minor cleanup
[ROCm/rccl commit: 90b2921207 ]
2019-11-01 13:30:01 -07:00
gilbertlee-amd
f9ef1553aa
Single Sync Timing mode ( #144 )
...
* Adding single sync timing mode to emulate timing reported by rccl-prim-test / rccl-tests
* Adding duration / overhead info
[ROCm/rccl commit: 2f9edd2432 ]
2019-11-01 10:18:25 -06:00
Jeff Daily
e43e1f1b3d
additional check for fine grain support in p2pCanConnect ( #146 )
...
[ROCm/rccl commit: 5a502955c9 ]
2019-10-31 08:58:38 -07:00
Wenkai Du
91b906cf88
rccl-prim-test: use hipExtLaunchMultiKernelMultiDevice and minor cleanup
...
[ROCm/rccl commit: ab91cdd5c9 ]
2019-10-30 13:15:02 -07:00
Gilbert Lee
a99accb2cb
Adding ability to switch between fine/coarse grain destination GPU memory
...
Adding ability to switch between memset/memcpy
[ROCm/rccl commit: 648c1ee7cc ]
2019-10-29 12:00:32 -06:00
Wenkai Du
b4ab922f94
Merge pull request #140 from scchan/rocm210_hc_function_calls
...
add -hc-function-calls switch back for HCC ROCm 2.10
[ROCm/rccl commit: 9be7ae8f0d ]
2019-10-28 09:56:47 -07:00
Michael LIAO
4b94f25d08
[cmake] Allow GPU targets to be parameterized with AMDGPU_TARGETS.
...
[ROCm/rccl commit: ec10a5cf14 ]
2019-10-25 13:55:27 -04:00
Wenkai Du
d3f399f619
Disable HDP flush for RDMA
...
[ROCm/rccl commit: 296176a4fd ]
2019-10-23 14:40:17 -07:00
Siu Chi Chan
8d2018d372
add -hc-function-calls switch back for HCC ROCm 2.10
...
[ROCm/rccl commit: d779eae1d0 ]
2019-10-21 18:00:02 -04:00
Wenkai Du
21bc1ef493
Revert collective chunk and slice steps to avoid drop in throughput
...
[ROCm/rccl commit: df74d12946 ]
2019-10-18 12:54:00 -07:00
saadrahim
0f4d4d63ec
CI Re-enabled for Ubuntu ( #135 )
...
[ROCm/rccl commit: a95529a6e2 ]
2019-10-18 11:38:51 -06:00
Gilbert Lee
7560929bd7
Reverting GenericOp bug workaround modifications to slice/chunk steps
...
[ROCm/rccl commit: 37603ae6cb ]
2019-10-11 09:20:10 -07:00
Gilbert Lee
cf597ff257
Performing __threadfence_system() with only first thread
...
[ROCm/rccl commit: 1392dd2997 ]
2019-10-11 09:16:19 -07:00
Gilbert Lee
d257970ad1
Fix for GenericOp device primitive bug
...
[ROCm/rccl commit: 8ae1bce3bb ]
2019-10-10 22:39:45 -07:00
Wenkai Du
fbcdfd8348
Merge pull request #136 from wenkaidu/tree
...
Enable tree kernels in build
[ROCm/rccl commit: 062c798c86 ]
2019-10-09 10:58:52 -07:00
Wenkai Du
f86ee41415
Enable tree kernels in build
...
Need to tune and specify NCCL_TREE_THRESHOLD to allow usage
[ROCm/rccl commit: 76976c9e2e ]
2019-10-08 23:20:11 +00:00
Changpeng Fang
d8a06589c9
Tuning the inline and unroll to reduce the scratch usage
...
Summary:
1. remove the noinline attribute for AllReduceThreeKernel;
2. change AUTPUNROLL for tree functions to 1 or 2;
Combining 1 and 2 will reduce the scratch usage from 1256 to 952
[ROCm/rccl commit: eec319038e ]
2019-10-08 14:02:25 -07:00
Siu Chi Chan
0af7f5268f
detect the hcc version and conditionally add the -hc-function-calls switch
...
[ROCm/rccl commit: b87ef4f152 ]
2019-10-03 13:25:25 -04:00
Wenkai Du
57dcac6afe
Only generate kernels for sum and copy
...
[ROCm/rccl commit: 61ef1dcad5 ]
2019-09-24 17:01:12 -07:00
Gilbert Lee
a401a91fd7
Re-adding gfx908 target
...
[ROCm/rccl commit: 6232985e34 ]
2019-09-13 16:57:34 +00:00
Gilbert Lee
dbb7500fd1
RDMA HDP flush fix
...
[ROCm/rccl commit: 86ce0a93b5 ]
2019-09-06 16:35:55 +00:00
Gilbert Lee
2847ad9576
Revert "Set RDMA default to off state"
...
This reverts commit 4afd6818ba .
[ROCm/rccl commit: 3e6b326a19 ]
2019-09-05 18:16:53 +00:00
rohit pathania
31670133a1
Read operation throughput
...
[ROCm/rccl commit: a270ee080e ]
2019-09-03 14:58:40 +05:30
rohit pathania
bc51b5bc28
display each workgroup ,links and directions with throughputs
...
[ROCm/rccl commit: e5b13d69e5 ]
2019-08-30 13:28:23 +05:30
Wenkai Du
daf2c4b200
Allocate opCount in pinned host memory for P2P transport
...
To avoid remote P2P read access when checking remote GPU's opCount
[ROCm/rccl commit: 8c975353ed ]
2019-08-29 10:22:09 -07:00
amdkila
ea0ce5c064
Merge pull request #128 from amdkila/hip-clang
...
Added hip-clang options to install script, and openmp/pthread flags
[ROCm/rccl commit: 259583cde6 ]
2019-08-27 16:23:40 -06:00
Wenkai Du
4afd6818ba
Set RDMA default to off state
...
[ROCm/rccl commit: 0f16ad966a ]
2019-08-26 10:59:33 -07:00
saadrahim
e433b21b23
Updating versioning to follow rocm-cmake standard ( #126 )
...
[ROCm/rccl commit: 544d4fb704 ]
2019-08-23 16:33:38 -06:00
Akila Premachandra
94b33a7550
Added hip-clang options to install script, and openmp/pthread options to CMakeLists.txt
...
[ROCm/rccl commit: f48ae5c98d ]
2019-08-23 22:02:42 +00:00
Wenkai Du
4df1defc3b
Merge pull request #125 from wenkaidu/fix_nvml_id
...
Assign unused nmvlDev to avoid random number
[ROCm/rccl commit: 6759660529 ]
2019-08-19 09:08:13 -07:00
Wenkai Du
54608abf5c
Merge pull request #117 from rpathani/xgmi_bench
...
Modified the code to use RTC clock frequency based on gpu gcn id
[ROCm/rccl commit: ee5dec4467 ]
2019-08-19 08:59:34 -07:00
rpathani
c441f2ff9b
Update rccl_prim_test.cpp
...
[ROCm/rccl commit: 40e30b5168 ]
2019-08-19 12:44:11 +05:30
Wenkai Du
04cd446d89
Assign unused nmvlDev to avoid random number
...
[ROCm/rccl commit: 86efdfc3b5 ]
2019-08-16 16:34:14 -07:00
Wenkai Du
60989a3fc9
Merge remote-tracking branch 'remotes/nccl/master' into HEAD
...
[ROCm/rccl commit: 7c38da0939 ]
2019-08-16 16:13:34 -07:00
Wenkai Du
7396d5c3ba
Tune AUTOUNROLL for better performance
...
Also remove all unused UNROLL defines
[ROCm/rccl commit: 1faededc03 ]
2019-08-16 10:34:53 -07:00