AI/rocm-systems
2
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
1,142 Commit 3 Branch 0 Tag
7e2d9053763c8fe9e2cb1ae8370fbc1dd4bb31be
Commit Graph

1142 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Wenkai Du aebed537a5 Reduce linking time through more parallel jobs (#657) 2022-11-30 16:06:03 -08:00
Wenkai Du fb9938cffa Query DMABuf support through HSA runtime API (#654) 2022-11-30 08:53:03 -08:00
Sylvain Jeaugey 28189e2df8 2.16.2-1 
Add support for CUDA 12.0, drop Kepler (sm_35).
Support for H100 features.
Make socket code more robust and protected. Solves #555.
Improve performance on large CUDA graphs, reducing dependencies.
Reduce inter-socket bandwidth on AMD CPUs to favor better paths.
Various fixes to ncclCommAbort.
Make service thread polling resistant to EINTR.
Compile with profiling API by default.
Extend NVTX instrumentation with call arguments.
 2022-11-30 02:31:59 -08:00
Wenkai Du 9594bbee3b Adjust P2P channels on Intel platform (#653) 2022-11-29 13:57:10 -08:00
akolliasAMD 11862f67de removed cmake HIP_CLANG_PATCH_LEVEL check (#652) 
* removed HIP_CLANG_PATCH_LEVEL check
 2022-11-29 09:48:59 -07:00
Wenkai Du 67d9327f52 Merge pull request #651 from wenkaidu/nccl_sync 
Sync up with NCCL
 2022-11-28 17:33:59 -08:00
Wenkai Du bf03a48289 Merge remote-tracking branch 'nccl/master' into HEAD 2022-11-28 09:46:16 -08:00
gilbertlee-amd 36ac8107bd Update CHANGELOG up to ROCm 5.4 (#649) 
* Update CHANGELOG for ROCm 5.4.0
 2022-11-23 09:40:19 -07:00
Sylvain Jeaugey 614b49f0de Fix google-fastsocket plugin build 2022-11-22 02:13:13 -08:00
Sylvain Jeaugey 55b1d8ab98 Add documentation for NCCL NET plugins 
Also repurpose dummy plugin as example, including headers and
compat layers from v6 to v2.
 2022-11-22 02:12:53 -08:00
Wenkai Du 57764f8152 Fix incorrect rocm-smi ID conversion (#648) 2022-11-21 19:44:39 -08:00
Wenkai Du 9cb72a3d0f Fix collective trace timestamp format (#647) 2022-11-21 08:11:12 -08:00
Wenkai Du cf3c32a626 Fix typo in previous hipify change (#645) 2022-11-15 11:51:47 -08:00
Wenkai Du b4f6eee9b4 Merge pull request #643 from ROCmSoftwarePlatform/2.15.5 
Sync up with NCCL 2.15.5
 2022-11-15 08:40:59 -08:00
Wenkai Du 562dd87036 Move hipify to cmake stage 
Add minimal ROCm/HIP version requirements for Graph support
 2022-11-14 18:10:45 +00:00
raramakr ca05b3d8d4 Merge pull request #644 from raramakr/swdev-reorg 
Correct hsa header path for new directory layout
 2022-11-10 10:06:39 -08:00
Wenkai Du 94ad7f6f51 Update tuning table and fix topo_expl 2022-11-07 18:24:24 +00:00
Ranjith Ramakrishnan b397cb16ea Correct hsa header path for new directory layout 2022-11-04 09:52:16 -07:00
Wenkai Du 9a077e6947 Merge remote-tracking branch 'nccl/master' into develop 2022-11-03 21:17:42 +00:00
Wenkai Du effc4b255b Merge pull request #641 from ROCmSoftwarePlatform/2.14.3 
Sync up with NCCL 2.14.3
 2022-11-02 08:31:12 -07:00
Wenkai Du 72ef100050 Fix P2P scheduling 2022-10-31 08:54:34 -07:00
Sylvain Jeaugey 2f4cb874ba Merge tag 'v2.15.5-1' 2022-10-25 01:15:22 -07:00
Sylvain Jeaugey cb111f764a 2.15.5-1 
Fix crash with CollnetChain on some node topologies
Fix hang when interleaving the capture of different graphs
Fix hang during init in multi-threaded mode
Fix potential data corruption with LL128 protocol on unaligned buffers.
Fix CPU usage during preconnect
Fixes double-free in the error path for ncclCommInitAll
Workaround hang on H100 with Ring/LL128 on 2 GPUs.
 2022-10-25 00:55:55 -07:00
Wenkai Du 4f0e223db4 Merge remote-tracking branch 'nccl/master' into develop 2022-10-20 15:41:29 +00:00
Wenkai Du bc8ef779df Fix missing initialization due to merge error (#640) 2022-10-19 21:20:11 -07:00
Wenkai Du fc554a2428 topo_expl: fix compilation error (#639) 2022-10-19 14:19:50 -07:00
gilbertlee-amd 10dbd2a452 Fixing formatting for copywrite (#638) 2022-10-19 13:43:21 -06:00
Wenkai Du 9ddf0e0649 Support P2P with invisible devices (#636) 
* Support P2P with invisible devices

* Update copyright year
 2022-10-17 10:24:59 -07:00
Wenkai Du 9916a09818 Merge pull request #634 from yzygitzh/ziyyang/npkit-fix 
Apply several fixes to NPKit
 2022-10-17 08:01:24 -07:00
raramakr b32f38126d Merge pull request #635 from raramakr/swdev 
Correct include and library path for new directory layout
 2022-10-14 15:48:44 -07:00
gilbertlee-amd ebb8b5bf63 Updating files for missing licenses (#637) 2022-10-14 13:49:16 -06:00
Ranjith Ramakrishnan cf4e963aaf Correct include and library path for new directory layout 
Use actual header files and libraries , rather than using wrapper header files and library softlinks
 2022-10-14 01:32:04 -07:00
Ziyue Yang 7d6bbc19d4 apply npkit 2022-10-14 01:28:17 +00:00
Sylvain Jeaugey d128d62238 Merge tag 'v2.15.1-1' 2022-10-07 11:00:26 -07:00
Edgar Gabriel 4972c129e3 Merge pull request #633 from edgargabriel/topic/topo-binary-tree 
introduce a hw topology aware bintree
 2022-10-05 17:06:54 -05:00
John Bachan 2401f4a918 Fixes a double-free in the error path of ncclCommInitAll. 
Fixes https://github.com/NVIDIA/nccl/issues/726
 2022-10-03 17:12:32 -07:00
Edgar Gabriel e645b02cd8 introduce a hw topology aware bintree 
for hayabusa architecture.
 2022-10-03 15:26:21 +00:00
gilbertlee-amd bd7d589446 Removing TransferBench from tools (#632) 
Point to new TransferBench repo
 2022-09-30 11:53:32 -06:00
akolliasAMD ef71550738 Added new gpu targets (#631) 2022-09-29 14:53:55 -06:00
Wenkai Du a523b37ac7 Another threadfence and flags rework (#629) 2022-09-28 16:49:29 -07:00
Wenkai Du 021932b3c8 Add LL128 tuning (#630) 2022-09-27 09:39:09 -07:00
Sylvain Jeaugey da8152e57a 2.15.1-1 
Add support for H100 (sm90).
Make sure NCCL kernel honor user stream priorities.
 2022-09-27 02:31:13 -07:00
Sylvain Jeaugey 99c28f2e75 Merge remote-tracking branch 'origin/master' 2022-09-27 02:24:41 -07:00
Cliff Woolley 78313a6d21 Use compatibility shim only with static cudart 
Closes issue 658
 2022-09-27 02:22:48 -07:00
Wen-Heng (Jack) Chung e8af0716c4 Merge pull request #619 from whchung/exp_reduce_code_size 
Only use split tree algorithm to reduce kernel code size.
 2022-09-26 10:06:27 -05:00
Sylvain Jeaugey ecab28a7c9 Fix potential deadlock during init in multi-thread mode. 
Make sure all calls calling cudaMalloc (including devCommSetup) are
called before the last bootstrapBarrier. That way, we avoid calls to
cudaMalloc be blocked by a NCCL kernel launched on another GPU by
another thread which completed init faster.

Resolve #623.
 2022-09-26 02:13:10 -07:00
Wen-Heng (Jack) Chung 35f1fe3434 Merge pull request #621 from whchung/exp_reduce_sleep_cycles 
Reduce s_sleep cycles
 2022-09-23 15:31:16 -05:00
Wen-Heng (Jack) Chung a80cc7e6e1 Merge pull request #624 from whchung/exp_tweak_unroll_factors 
Tweak unroll factors.
 2022-09-23 11:30:05 -05:00
Wen-Heng (Jack) Chung a08a24e042 Merge pull request #625 from whchung/exp_sync_lds 
Abolish syncthreads and only wait on LDS traffic.
 2022-09-23 11:29:48 -05:00
Wen-Heng (Jack) Chung 84054c3b30 Tweak unroll factors. 2022-09-22 13:03:04 -05:00