Mustafa Abduljabbar a075779dcd Use batched P2P to enhance alltoall small message performance (#1902) ...

* Batch P2P operations (2 per CU/channel) and update channel-part mapping

- Revert bitreversal and fix channel mapping to be compatible with P2P batching and avoid hangs

- P2P batching is only used for more than 2 nodes to avoid aggregating intra-node traffic when it is dominant for less than 2 nodes

* Address single node regression and channel per net peer

* Add batching threshold

* Add enable switch for batching

* Update CHANGELOG.md

* Add minor comment change

* Update CHANGELOG.md

Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>

* Update CHANGELOG.md

Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>

[ROCm/rccl commit: c1e1f2faeb]

2025-09-22 16:25:10 -04:00

projects/rccl

Use batched P2P to enhance alltoall small message performance (#1902)

2025-09-22 16:25:10 -04:00

S

Description

No description provided

282 MiB

Languages

C++ 67.5%

C 20.6%

Python 6.6%

CMake 3.4%

Shell 0.6%

Other 1.1%