systems-assistant[bot] 58c203e252 Fix channel overuse for 1 rank comms (#2760) ...

* Fix channel overuse for 1 rank comms

* limit channels when warpSpeed is enabled but not used

* enable std::min check against # of CUs for maxChannels computation when warpSpeed is enabled

---------

Co-authored-by: Mustafa Abduljabbar <muabdulj@amd.com>
Co-authored-by: isaki001 <ioannissakiotis@gmail.com>
Co-authored-by: Corey Derochie <161367113+corey-derochie-amd@users.noreply.github.com>

2026-01-29 12:13:46 -06:00

..

device

fix bug in reduce kernel bfloat16 for ROCm >= 6.0 (#2139)

2026-01-20 14:07:20 -08:00

graph

Fix channel overuse for 1 rank comms (#2760)

2026-01-29 12:13:46 -06:00

include

Fix channel overuse for 1 rank comms (#2760)

2026-01-29 12:13:46 -06:00

misc

[DEVICE] Switch to amd-smi from rocm-smi (#1759)

2026-01-21 09:05:47 -06:00

nccl_device

NCCL 2.28.3-1

2025-09-02 13:53:34 -07:00

plugin

Merge remote-tracking branch 'nccl/master' into develop

2026-01-20 13:04:02 -06:00

ras

Merge remote-tracking branch 'nccl/master' into develop

2026-01-20 13:04:02 -06:00

register

Merge remote-tracking branch 'nccl/master' into develop

2026-01-20 13:04:02 -06:00

scheduler

Merge remote-tracking branch 'nccl/master' into develop

2026-01-20 13:04:02 -06:00

transport

applying the changes from net_ib.cc to rocm_net_ib.cc to ensure DMABUF-disabled configurations are respected. (#2152)

2026-01-21 12:11:56 -08:00

allocator.cc

Merge remote-tracking branch 'nccl/master' into develop

2026-01-20 13:04:02 -06:00

bootstrap.cc

Merge remote-tracking branch 'nccl/master' into develop

2026-01-20 13:04:02 -06:00

ce_coll.cc

Merge remote-tracking branch 'nccl/master' into develop

2026-01-20 13:04:02 -06:00

channel.cc

NCCL 2.27.3-1

2025-05-29 20:56:40 -07:00

CMakeLists.txt

NCCL 2.28.3-1

2025-09-02 13:53:34 -07:00

collectives.cc

revert memcpy use for direct AG (#2146)

2026-01-20 13:58:28 -06:00

commDump.cc

Add ncclCommDump API (#2068)

2025-12-11 15:02:35 -07:00

debug.cc

Merge remote-tracking branch 'nccl/master' into develop

2026-01-20 13:04:02 -06:00

dev_runtime.cc

Merge remote-tracking branch 'nccl/master' into develop

2026-01-20 13:04:02 -06:00

enhcompat.cc

2.12.7-1

2022-03-02 20:48:56 +01:00

enqueue.cc

fix potential segfaults due to use after malloc fails (#2137)

2026-01-20 14:11:29 -08:00

group.cc

Merge remote-tracking branch 'nccl/master' into develop

2026-01-20 13:04:02 -06:00

init_nvtx.cc

Merge remote-tracking branch 'nccl/master' into develop

2026-01-20 13:04:02 -06:00

init.cc

Fix channel overuse for 1 rank comms (#2760)

2026-01-29 12:13:46 -06:00

Makefile

NCCL 2.28.3-1

2025-09-02 13:53:34 -07:00

mnnvl.cc

Merge remote-tracking branch 'nccl/master' into develop

2026-01-20 13:04:02 -06:00

msccl.cc

Deprecated MSCCL API functions (#1740)

2025-06-11 17:52:09 -06:00

nccl.h.in

use message instead of warning for nccl.h C++ check (#2128)

2026-01-20 14:21:38 -07:00

nccl.pc.in

Add pkgconfig file (#190)

2019-04-08 09:16:54 -07:00

proxy.cc

fix potential segfaults due to use after malloc fails (#2137)

2026-01-20 14:11:29 -08:00

rccl_wrap.cc

[DEVICE] Switch to amd-smi from rocm-smi (#1759)

2026-01-21 09:05:47 -06:00

sym_kernels.cc

Merge remote-tracking branch 'nccl/master' into develop

2026-01-20 13:04:02 -06:00

transport.cc

Merge remote-tracking branch 'nccl/master' into develop

2025-08-28 15:46:28 -05:00