Feat: sm90 support mxfp8 fp8 groupgemm#362
Open
zhangxiaolei123456 wants to merge 17 commits into
Open
Conversation
The MXFP8 consumer was issuing a fresh global load for every (kk, accum) pair, even though the producer warp had already staged SFA/SFB into shared memory. Replace the per-iteration GMEM lookups with two LDS loads (one 32-bit pack per SFA row, one 64-bit pack per pair of SFB rows) outside the wgmma kk loop, and decode the per-kk byte with a register-side shift. Keeps the result bit-identical while removing the GMEM dependency from the inner wgmma pipeline.
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
accuracy test:
python -m pytest tests/test_sm90_mxfp8_fp8.py -q OK
Peformance test: