kanshan/nccl-tests
1
0
Fork 0
mirror of https://github.com/NVIDIA/nccl-tests.git synced 2026-01-14 02:47:21 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

Updated with new command line arguments

Browse Source
This commit is contained in:
David Addison 2021-07-06 14:47:50 -07:00
parent f476f4a17a
commit 11cff17a04
1 changed files with 4 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
README.md
View File

@ -52,19 +52,21 @@ All tests support the same set of arguments :
* `-i,--stepbytes <increment size>` fixed increment between sizes. Default : (max-min)/10.
* `-f,--stepfactor <increment factor>` multiplication factor between sizes. Default : disabled.
* NCCL operations arguments
* `-o,--op <sum/prod/min/max/all>` Specify which reduction operation to perform. Only relevant for reduction operations like Allreduce, Reduce or ReduceScatter. Default : Sum.
* `-o,--op <sum/prod/min/max/avg/all>` Specify which reduction operation to perform. Only relevant for reduction operations like Allreduce, Reduce or ReduceScatter. Default : Sum.
* `-d,--datatype <nccltype/all>` Specify which datatype to use. Default : Float.
* `-r,--root <root/all>` Specify which root to use. Only for operations with a root like broadcast or reduce. Default : 0.
* Performance
* `-n,--iters <iteration count>` number of iterations. Default : 20.
* `-w,--warmup_iters <warmup iteration count>` number of warmup iterations (not timed). Default : 5.
* `-m,--agg_iters <aggregation count>` number of operations to aggregate together in each iteration. Default : 1.
* `-a,--average <0/1/2/3>` Report performance as an average across all ranks (MPI=1 only). <0=Rank0,1=Avg,2=Min,3=Max>. Default : 1.
* Test operation
* `-p,--parallel_init <0/1>` use threads to initialize NCCL in parallel. Default : 0.
* `-c,--check <0/1>` check correctness of results. This can be quite slow on large numbers of GPUs. Default : 1.
* `-z,--blocking <0/1>` Make NCCL collective blocking, i.e. have CPUs wait and sync after each collective. Default : 0.
* `-G,--cudagraph <num graph launches>` Capture iterations as a CUDA graph and then replay specified number of times. Default : 0.
## Copyright
NCCL tests are provided under the BSD license. All source code and accompanying documentation is copyright (c) 2016-2019, NVIDIA CORPORATION. All rights reserved.
NCCL tests are provided under the BSD license. All source code and accompanying documentation is copyright (c) 2016-2021, NVIDIA CORPORATION. All rights reserved.