AI/rocm-systems
2
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
Files
cf48e57bd943f970abdd63f06ee1d45b571ee474
rocm-systems/projects
T
History
corey-derochie-amd cf48e57bd9 Moved mscclpp_ncclGetUniqueId call into ncclCommInitRankFunc (#1332) 
* Moved call to `mscclpp_ncclGetUniqueId` into `ncclCommInitRankFunc` to avoid setting up transport early in environments where MSCCL++ isn't valid.

* Checking `mscclEnabled` for the process and the topology to gate MSCCL++.

* Allowed `mscclForceEnable` to enable MSCCL++.

[ROCm/rccl commit: 853a0586b4]
2024-09-16 16:41:40 -06:00
..
rccl
Moved mscclpp_ncclGetUniqueId call into ncclCommInitRankFunc (#1332)
2024-09-16 16:41:40 -06:00