Conversation
✅ All Checks Passed — Ready for Review
📖 Need help? See the Policy FAQ for details on every check and how to fix failures. |
|
🎉 All checks passed! This PR is ready for review. |
Codecov Report❌ Patch coverage is ❌ Your project check has failed because the head coverage (46.49%) is below the target coverage (80.00%). You can increase the head coverage or adjust the target coverage. Additional details and impacted files@@ Coverage Diff @@
## develop #12194 +/- ##
===========================================
- Coverage 69.68% 69.32% -0.36%
===========================================
Files 2803 2804 +1
Lines 461587 462043 +456
Branches 68031 68121 +90
===========================================
- Hits 321634 320305 -1329
- Misses 116678 118378 +1700
- Partials 23275 23360 +85
*This pull request uses carry forward flags. Click here to find out more.
🚀 New features to boost your workflow:
|
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Motivation
This PR extends the optional RCCL backend to MPI plans so the transpose can use RCCL collectives (GPU-direct, Slingshot/OFI when the net plugin is present) instead of MPI. The product rule is the same as a one-GPU-per-rank distributed FFT: RCCL is used only when every rank in the plan communicator owns exactly one GPU.
Technical Details
RCCL is used for the multi-process transpose only when every MPI rank owns exactly one GPU (NCCL rank = sorted (mpi_rank, device); with that gate, NCCL rank equals MPI rank). Otherwise, or with ROCFFT_RCCL_DISABLE=1, the existing MPI A2A/P2P path remains.
Rank 0 broadcasts ncclUniqueId, each rank calls ncclCommInitRank for its local device, and CommRCCLAllToAll launches only local buffers. MPI is still used for brick metadata and this bootstrap.
Test Plan
Test locally on available AMD clusters and by CI,
Test Result
Local tests passed. CI ongoing.
Submission Checklist
JIRA ID: AIFFT-456