058886cb20
Add another Rome model ( #1095 )
...
[ROCm/rccl commit: df98a6957d
2024-02-28 10:46:05 -08:00
cee279fd99
Implement ROCTX ( #1094 )
...
* Implement roctx
[ROCm/rccl commit: b617aecc31
2024-02-27 15:46:15 -07:00
5fa796af0c
2.20.5-1
...
Fix UDS connection failure when using ncclCommSplit. Issue #1185
[ROCm/rccl commit: 48bb7fec79
2024-02-26 02:52:39 -08:00
874998033f
Add new GPU model ( #1080 )
...
[ROCm/rccl commit: 74f9e5db64
2024-02-23 12:19:42 -08:00
4b31894d70
Update RCCL/MSCCL work FIFO depth to 256K ( #1091 )
...
[ROCm/rccl commit: c5ab37211b
2024-02-21 17:15:11 -08:00
4b4bdd904e
LL128 check if all XGMI ( #1089 )
...
[ROCm/rccl commit: b275ed0b56
2024-02-21 09:41:40 -07:00
32e1c8cba0
Merge pull request #1079 from BertanDogancay/2.19.4-sync
...
2.19.4 Sync
[ROCm/rccl commit: 2fb12a9358
2024-02-16 09:50:11 -07:00
e0dd21028f
Allow bus id to be null ( #1085 )
...
* Allow bus id to be null
[ROCm/rccl commit: bac57421c7
2024-02-15 16:36:51 -07:00
ef72944015
Disable unsupported ld/st instructions
...
[ROCm/rccl commit: 6f3310605c
2024-02-15 13:58:16 -08:00
7842411fb3
Merge remote-tracking branch 'rccl/develop' into 2.19.4
...
[ROCm/rccl commit: 76f83f95ab
2024-02-15 13:37:14 -08:00
c4e9e2b18a
Use native half without conversion ( #1083 )
...
[ROCm/rccl commit: 51003c9980
2024-02-13 16:57:34 -08:00
2f14acf770
Fix undefined symbol when nvtx is not enabled ( #1082 )
...
[ROCm/rccl commit: 1f0af90206
2024-02-13 14:03:43 -08:00
2ab8a3a750
2.20.3-1
...
Add support for alternating rings, allow for cross-nic rings without
cross-rail communication.
Add support for user buffer registration for network send/recv.
Optimize aggregated operations to better utilize all channels.
Add flattening for BCM PCI gen5 switches.
Add support for inter-node NVLink communication
Add support for port fusion in NET/IB.
Add support for ReduceScatter and AllGather using Collnet.
Update net API to v8.
Fix hang during A2A connection.
[ROCm/rccl commit: b6475625fb
2024-02-13 04:22:38 -08:00
de6f20b7ae
Fix docs
...
[ROCm/rccl commit: 32cca51894
2024-02-11 22:32:55 -08:00
d5f5091e5d
Merge remote-tracking branch 'rccl/develop' into 2.19.4
...
[ROCm/rccl commit: d999d9ad21
2024-02-09 11:31:03 -06:00
6775a75906
2.18.5 fix ( #1077 )
...
* Revert "Revert "2.18.5-1""
This reverts commit 7cc572ecf95669b0d7b6
2024-02-09 09:18:38 -08:00
45ed3ef4e7
Nvtx support ( #1076 )
...
* NVTX support
[ROCm/rccl commit: 8a442faa12
2024-02-08 14:08:24 -07:00
1538b908ac
msccl: use relaxed atomics on scratch buffer ( #1075 )
...
[ROCm/rccl commit: 5257c753c5
2024-02-08 12:09:56 -08:00
ce39eefe65
Doubling P2P channels per peer on single node gfx94x only ( #1074 )
...
[ROCm/rccl commit: 704c9ef0d1
2024-02-07 14:05:57 -08:00
57e508f2e4
Doubling P2P channels per peer on single node only ( #1069 )
...
[ROCm/rccl commit: 1d989f6524
2024-02-02 12:41:00 -08:00
b17f5947ba
Revert re-usage of connect and listen ports
...
[ROCm/rccl commit: 12ac20ade5
2024-02-01 10:03:13 -08:00
ef57c82b60
Clean up
...
[ROCm/rccl commit: 00fdb1ef51
2024-01-31 17:27:15 -08:00
5ce968f570
Fix stack size
...
[ROCm/rccl commit: da85abab54
2024-01-31 17:09:07 -08:00
fad3417dc1
Fix transport merge
...
[ROCm/rccl commit: 95f87232c4
2024-01-31 17:35:12 -06:00
4f4e52aafe
Merge remote-tracking branch 'rccl/develop' into 2.19.4
...
[ROCm/rccl commit: 1a134b283b
2024-01-31 11:53:10 -06:00
c2c9ed2acb
Merge remote-tracking branch 'nccl/master' into develop
...
[ROCm/rccl commit: 9ff53eeeae
2024-01-30 14:43:43 -08:00
96479caeda
Include common.h in enqueue.cc instead ( #1067 )
...
[ROCm/rccl commit: 01b359027b
2024-01-30 08:24:22 -08:00
3e66d42d95
msccl: ensure memory coherence after data receive ( #1062 )
...
[ROCm/rccl commit: f7550d83b8
2024-01-30 08:22:50 -08:00
a2b3e1ab2d
correct data type
...
[ROCm/rccl commit: 31ec5d5cb0
2024-01-28 19:55:19 -08:00
a184aabda5
modifying the tuning table to improve the performance of allreduce for 8MB and 16MB for single-node MI300X ( #1063 )
...
[ROCm/rccl commit: ccfb35fa6d
2024-01-26 09:05:53 -05:00
cbc6aa13e7
colltrace: fix dropped trace messages ( #1059 )
...
* colltrace: fix dropped trace messages
* Remove extra space
[ROCm/rccl commit: be8ef4367f
2024-01-25 13:31:53 -08:00
71a356670f
Increase P2P channels per peer ( #1060 )
...
[ROCm/rccl commit: ffde530af5
2024-01-25 11:21:58 -08:00
a7a60a06ec
Fix sendrecv merge
...
[ROCm/rccl commit: 4aafb2a3c5
2024-01-24 16:23:53 -08:00
404d398bac
Merge remote-tracking branch 'nccl/v2.19' into develop
...
[ROCm/rccl commit: 81ddf9de89
2024-01-24 15:25:33 -08:00
17f70df058
Revert "msccl: build same number of kernels as in ROCm 5.7" ( #1058 )
...
This reverts commit f960174d03be7e5174baa83b256526d388a38842.
[ROCm/rccl commit: 7987015a19
2024-01-24 08:43:50 -08:00
645411f8f4
Use binary search for direct function calls ( #1057 )
...
* Use binary search for direct function calls
* fix scratch mem issue on MI300
[ROCm/rccl commit: 5564d65e71
2024-01-22 17:37:56 -07:00
56482a8be8
Fix collective trace when rccl is configured ( #1056 )
...
* Fix collective trace when rccl is configured
[ROCm/rccl commit: c4dbf8a914
2024-01-22 09:26:44 -07:00
8b8179a689
Use new HIP graph API compatible with CUDA 11030 ( #991 )
...
* Use new HIP graph API compatible with CUDA 11030
* Update dependency to ROCm 6.1
* Fix single stream use case
[ROCm/rccl commit: 7e25d5bc55
2024-01-21 19:00:50 -08:00
1c9b4fab39
COLLECTIVES: Switch to unroll 2 for MI300 ( #1051 )
...
Signed-off-by: nileshnegi <Nilesh.Negi@amd.com >
[ROCm/rccl commit: 8b97a20943
2024-01-19 12:16:05 -06:00
11674674fc
[DEV] Configure functions in RCCL ( #986 )
...
* configure functions in rccl
[ROCm/rccl commit: 28d9b170c9
2024-01-18 15:07:16 -07:00
dbf906d8fa
Only use full MAXCHANNELS for gfx94x ( #1050 )
...
[ROCm/rccl commit: 3325f96c56
2024-01-17 09:00:49 -08:00
5c349cd729
adding rccl tuning parameters for MI300X gfx942 with 8 GPUs single and multi-node ( #1047 )
...
[ROCm/rccl commit: b08124c85d
2024-01-16 13:44:32 -05:00
8d38747c65
Add option to force enable network transport on single node ( #1046 )
...
[ROCm/rccl commit: 261707d90a
2024-01-16 07:54:18 -08:00
7cc572ecf9
Revert "2.18.5-1"
...
This reverts commit 26b91b9dbb767fde8210
2024-01-12 16:54:19 +00:00
a056463d4d
Addressing the compiler warning ( #988 )
...
[ROCm/rccl commit: cf248d9402
2024-01-10 14:59:40 -07:00
df0dbb887d
cover more gpu/nic mapping cases ( #1037 )
...
[ROCm/rccl commit: 735178c1fe
2024-01-10 08:01:37 -08:00
64cf812da0
Re-enable L128 on gfx90a of compiler supports it ( #1036 )
...
[ROCm/rccl commit: 5851ae5974
2024-01-10 08:01:11 -08:00
cec06d59d1
Un-escaped character causes error with address sanitizer builds ( #992 )
...
Signed-off-by: Nilesh M Negi <Nilesh.Negi@amd.com >
Co-authored-by: Jenkins <jenkins-compute@amd.com >
[ROCm/rccl commit: 249e9f7f65
2024-01-09 13:28:32 -06:00
d3a47bb387
Merge pull request #1022 from PedramAlizadeh/sync_nccl_2.18.6
...
Sync to nccl 2.18.6
[ROCm/rccl commit: aa5c84c997
2024-01-09 13:29:29 -05:00
30f744dc35
msccl: use custom reduce function ( #1033 )
...
[ROCm/rccl commit: d9871d171b
2024-01-08 14:53:12 -08:00
Wenkai Du
Bertan Dogancay
Sylvain Jeaugey
akolliasAMD
BertanDogancay
Wenkai Du
Pedram Alizadeh
Nilesh M Negi
Hossein Pourreza