kanshan/nccl-tests
1
0
Fork 0
mirror of https://github.com/NVIDIA/nccl-tests.git synced 2026-04-17 13:08:18 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

Commit Graph

  • ab2c8576e3
         Merge 56d0a64f8b into f727aa2a54 kzlxd 2026-04-17 11:34:52 +0800
  • dac56a51bf
         Merge 654abc2246 into f727aa2a54 evelyn2908 2026-04-13 13:41:11 -0700
  • f727aa2a54 NCCL_TESTS_VERSION 2.18.3 master v2.18.3 David Addison 2026-04-13 13:37:21 -0700
  • 5dc0670e56 Add maxP2pPeers comm config for sendrecv David Addison 2026-03-30 11:39:31 -0700
  • 56d0a64f8b Add multi-channel DP test results kangzhen02 2026-04-01 09:57:40 +0800
  • 2aca431208
         Merge bfbccd060a into af1dcac92a Stas Bekman 2026-03-13 08:30:32 +0000
  • 51c5899c85
         Merge cac16990b8 into af1dcac92a yangxingwu 2026-03-13 07:42:43 +0000
  • d4deb24751
         Merge 0890dff640 into af1dcac92a Jun Jiang 2026-03-13 05:45:35 +0000
  • af1dcac92a NCCL_TESTS_VERSION 2.18.2 v2.18.2 David Addison 2026-03-11 15:35:13 -0700
  • eb0d3d2a00 Display unalign setting in output David Addison 2026-03-11 15:05:54 -0700
  • e02c20b898 NCCL_TESTS_VERSION 2.18.1 v2.18.1 David Addison 2026-03-11 09:55:31 -0700
  • c1af7df1f3 Update -z option description in README.md David Addison 2026-03-11 09:47:10 -0700
  • ba52a70492 Allow blocking collectives without MPI_Barrier in timing loop David Addison 2026-03-11 09:26:08 -0700
  • 8d26b23319 Allocate buffers during thread initialization Theofilos Ioannis Manitaras 2026-03-10 18:33:50 +0100
  • dd0bafd178 NCCL_TESTS_VERSION 2.18.0 v2.18.0 David Addison 2026-03-06 17:55:12 -0800
  • 115fb09377 Add new unalign flag to README.md and update help text David Addison 2026-03-06 17:45:03 -0800
  • e986a6156c Add -u <index> to force unaligned buffer addresses David Addison 2026-03-06 17:30:17 -0800
  • c379e19a71 NCCL_TESTS_VERSION 2.17.10 v2.17.10 David Addison 2026-03-05 15:35:13 -0800
  • db221defdb Request GIN to be explicitly enabled in all to all test Ahsan Pervaiz 2026-03-03 14:05:52 -0800
  • ae98985f55 Fix Clang compilation errors with VLA initialization Marcin Malagowski 2026-02-09 15:10:55 +0200
  • d7451ee3f0 Fix Clang compilation errors with VLA initialization Marcin Malagowski 2026-02-09 15:10:55 +0200
  • 9938d5a657 Fix compilation issues with latest NCCL release headers David Addison 2026-02-04 16:43:20 -0800
  • 2535da805b NCCL_TESTS_VERSION 2.17.9 v2.17.9 David Addison 2026-02-03 11:04:48 -0800
  • 85ca91d1b1 Fix: corrected typos in the JSON output mykeduong 2026-02-02 18:43:18 -0800
  • 1442aaa1c0 Fix: corrected typos in the JSON output mykeduong 2026-02-02 18:43:18 -0800
  • a1b915f5d0 fix nccl version for cuda13x xlliu 2026-01-28 09:00:09 +0000
  • 63cb76d463 fix allow-change-held-packages for cuda12.9 Dockerfile xlliu 2026-01-28 07:46:52 +0000
  • 613b80389c add Dockerfile for cuda129 xlliu 2026-01-28 07:12:41 +0000
  • 88d7e33207 Add -M memory report option to README.md David Addison 2026-01-15 13:32:55 -0800
  • 81463c58d0 NCCL_TESTS_VERSION 2.17.8 v2.17.8 David Addison 2026-01-06 15:00:17 -0800
  • 7278698c1b Clarified use of Mebibytes and Gibibytes for sizes David Addison 2026-01-06 14:59:17 -0800
  • 2656c58421 NCCL_TESTS_VERSION 2.17.7 v2.17.7 Katie Gioioso 2025-12-30 20:16:39 +0000
  • 070d17528c refactor comm init Katie Gioioso 2025-11-21 02:56:37 +0000
  • 332e61896f device api 2.28 is not compatible with 2.29. Check versions and print error if there is a mismatch Katie Gioioso 2025-11-25 19:11:18 +0000
  • 24874bdaa8 Compatibility with 2.29 device API: use NCCL_DEV_COMM_REQUIREMENTS_INTIIALIZER, query properties to check for device api support Katie Gioioso 2025-11-13 01:16:07 +0000
  • 7106245178 Add include of <limits> due to compilation error David Addison 2025-11-17 15:50:15 -0800
  • 760c467f12 Add memory usage report option David Addison 2025-11-02 20:32:20 -0800
  • 587d57c5ae update release.yml xlliu-scitix 2025-12-21 16:05:17 +0000
  • db45ae9c3d update dockerfile xlliu-scitix 2025-12-21 15:40:24 +0000
  • 9b37f54680 update dockerfile xlliu-scitix 2025-12-21 15:11:53 +0000
  • ea941cea48 update dockerfile xlliu-scitix 2025-12-21 12:11:36 +0000
  • 0b7aeaf09b update dockerfile xlliu-scitix 2025-12-21 12:05:00 +0000
  • a63433ce70 update dockerfile xlliu-scitix 2025-12-21 12:00:18 +0000
  • 2225d3f05e update dockefile xlliu-scitix 2025-12-21 10:56:28 +0000
  • eb04f2127a update dockerfile xlliu-scitix 2025-12-21 10:48:36 +0000
  • d5d08b99cc update Dockerfile xlliu-scitix 2025-12-21 10:00:50 +0000
  • e115bd8517 fix bugs xlliu-scitix 2025-12-21 08:39:06 +0000
  • c9ace03f2d update pre-check.yml xlliu-scitix 2025-12-21 08:12:47 +0000
  • 5fe638cb66 update Dockerfile xlliu-scitix 2025-12-21 08:02:26 +0000
  • b5c25a812c delete duplicated pre-check.yml xlliu-scitix 2025-12-21 07:49:52 +0000
  • e9e9a27320 Merge branch 'sicl' into sync/upstream-20251216 xlliu-scitix 2025-12-21 07:41:39 +0000
  • 42c1025d7d add free disk space in release.yml ci xlliu-scitix 2025-12-21 07:18:22 +0000
  • 5eec99dc08
         add dockerfile and ci workflow (#1) xlliu-scitix 2025-12-21 15:12:18 +0800
  • cc230e61e1 add github workflow xlliu 2025-12-21 14:13:42 +0800
  • a7444d2ed8 chore(scripts): add install scripts xlliu-scitix 2025-12-21 03:12:56 +0000
  • 7dce185a16 update Dockerfile xlliu 2025-12-21 10:18:09 +0800
  • 4aebbcbf93 add Dockerfile xlliu 2025-12-21 10:15:30 +0800
  • a922a05e60 update scripts 刘小丽 2025-12-19 20:13:29 +0800
  • 351b3c1fd9 add scripts 刘小丽 2025-12-19 10:05:04 +0800
  • 8ab1ad2779 Merge remote-tracking branch 'upstream/master' into sync/upstream-20251216 刘小丽 2025-12-16 23:05:46 +0800
  • 654abc2246
         Create azure-webapps-node.yml evelyn2908 2025-11-22 09:36:01 +0800
  • 4bc314aa27 Add README.md text for -J option David Addison 2025-11-21 11:31:48 -0800
  • 51f2e7ed7c Remove trailing WS when timestamp option not used David Addison 2025-11-03 11:18:39 -0800
  • da0b547b1b NCCL_TESTS_VERSION 2.17.6 v2.17.6 David Addison 2025-10-28 10:20:47 -0700
  • e2af90af76 Add new report_timestamps option to README.md David Addison 2025-10-28 10:18:53 -0700
  • a62c975681 Add option to suffix a timestamp to each perf line David Addison 2025-10-28 09:58:21 -0700
  • 0bb567cc02 NCCL_TESTS_VERSION 2.17.5 v2.17.5 David Addison 2025-10-28 09:34:56 -0700
  • 013c49e930 add necessary ifdef guards for device API tests Shane Snyder 2025-10-27 10:05:24 -0700
  • f66d20e360 add runtime guards for ncclAlltoAll() Shane Snyder 2025-10-27 13:38:10 -0700
  • 3744121a2d NCCL_TESTS_VERSION 2.17.4 v2.17.4 David Addison 2025-10-24 16:36:21 -0700
  • 9641693e9b Add PRINT of nccl-tests, NCCL header and library versions David Addison 2025-10-24 15:57:59 -0700
  • 9829ea42b5 add GIN-based device API kernels to alltoall Shane Snyder 2025-10-06 08:49:00 -0500
  • e483d59a8f test on conda Yusuo Hu 2025-10-22 11:32:36 -0700
  • 00f52811b8 Add support for JSON output to perf test framework v2.17.3 David Addison 2025-10-17 12:01:25 -0700
  • c8e5575f50
         Add Option to Print Timestamps for Infinite Cycles (#1) yakovdyadkin 2025-10-14 10:48:16 -0500
  • 438f4d5dfc
         Merge fd03a19727 into abc46770a9 dmitrygx 2025-10-08 16:28:41 -0700
  • abc46770a9 Check if sufficient GPUs are available v2.17.2 Stephen Sachs 2025-09-26 14:34:23 +0200
  • 0890dff640 Use Family-Specific architecture features on CUDA 13, enable Thor and Spark support jasl 2025-09-17 17:48:35 +0800
  • 9a5c15461a Fix compilation for old NCCL versions v2.17.1 Sylvain Jeaugey 2025-09-05 01:10:57 -0700
  • e12dbb0a14 Update to align with the NCCL 2.28 release v2.17.0 David Addison 2025-09-04 17:23:22 -0700
  • c2cb96faac Update NVCUFLAGS and CXXFLAGS to use -std=c++14 v2.16.9 David Addison 2025-08-29 14:55:31 -0700
  • f2015cbe82 Modified warmup to run for more message sizes v2.16.8 David Addison 2025-08-25 13:47:54 -0700
  • fae7cb4727
         Merge pull request #316 from martin-belanger/print-program-name v2.16.7 David Addison 2025-07-24 14:58:54 -0700
  • 6edafa0a9c Add extra reserved space during maxBytes calculation v2.16.6 David Addison 2025-05-21 09:40:26 -0700
  • def2d3689c Minor fix to Makefile v2.16.5 David Addison 2025-07-23 16:04:30 -0700
  • 5013240b74 json output support zhiyongww 2025-06-09 11:58:51 -0700
  • 97ee098516 Add Turing (SM75) support to CUDA 13.0 builds v2.16.4 David Addison 2025-06-04 17:54:58 -0700
  • e7c8825b0b Wrap ncclCommWindowRegister() calls within ncclGroup v2.16.3 David Addison 2025-06-03 10:36:53 -0700
  • dafb70408d Print the name of the program being executed Martin Belanger 2025-06-03 11:43:02 -0400
  • 5290298ab6 Reinstate Pascal suppport for CUDA 12.8+ builds v2.16.2 David Addison 2025-06-02 09:29:52 -0700
  • 8bc16f4e01 Need to drop Volta (sm_70) support from CUDA 13.0 v2.16.1 David Addison 2025-05-30 18:04:25 -0700
  • 0c60e6a8e4 Fix formatting errors in README.md David Addison 2025-05-30 17:43:30 -0700
  • a5c539e68b Add support for Symmetric Memory Registration David Addison 2025-05-19 18:20:22 -0700
  • 9b9de7da81 Add support for Symmetric Memory Registration v2.16.0 David Addison 2025-05-19 18:20:22 -0700
  • 598fc0060e Add extra reserved space during maxBytes calculation delete David Addison 2025-05-21 09:40:26 -0700
  • e041d901e6 Re-add sm_70 support for CUDA 12.8+ and 13.0 builds v2.15.2 David Addison 2025-05-07 10:30:59 -0700
  • 1021260ca9 Make verifiable a DSO and add NAME_SUFFIX support v2.15.1 David Addison 2025-04-21 11:26:35 -0700
  • 44f2be9989
         Merge 2803191af6 into 501a149d57 x41lakazam 2025-04-22 09:23:04 -0700
  • 501a149d57 Add support for FP8 datatypes v2.15.0 David Addison 2025-04-18 19:20:59 -0700
  • 888e412525 feat: add a shell script to record nccl-test result zhushuang 2025-04-09 18:56:36 +0800