[QST]What does Epilogue do?(Especially in Hopper)
Nobody has claimed this yet.
- Dominant language
- C++
- Stars
- 10.5k
- Forks
- 2.1k
- Avg merge
- 3d 11h
- Merged PRs (30d)
- 7
Description
I am writing example 48, and I noticed file:
https://github.com/NVIDIA/cutlass/blob/06b21349bcf6ddf6a1686a47a137ad1446579db9/include/cutlass/gemm/kernel/sm90_gemm_tma_warpspecialized_cooperative.hpp#L321
I find out with surprise that, epilogue is in producer, but we do not even enter epilogue?! Previously I guess we save result using epilogue, both rearrange it into shared memory and then save into global, but now, storing is done in cousumer...?
Contributor guide
No contributing guide indexed for this repository
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start with include/cutlass/gemm/kernel/sm90_gemm_tma_warpspecialized_cooperative.hpp around line 321 and compare it with the context of example 48. Trace how the producer, consumer, and epilogue roles relate to storing results in Hopper. Done means providing a clear explanation of what the epilogue does and why the observed execution path places storing in the consumer.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- cpp
- Domain
- hpc, performance
- Issue type
- Documentation
- Difficulty
- 4/5
- Estimated time
- 1-2 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 35/100