Relax MXFP8 GEMM K constraint from multiple-of-128 to multiple-of-64#642
Open
JohnQinAMD wants to merge 1 commit into
Open
Relax MXFP8 GEMM K constraint from multiple-of-128 to multiple-of-64#642JohnQinAMD wants to merge 1 commit into
JohnQinAMD wants to merge 1 commit into
Conversation
JohnQinAMD
requested review from
ipanfilo,
wangye805 and
wenchenvincent
as code owners
The ROCm MXFP8 GEMM guard required K % 128 == 0, but hipBLASLt has MXFP8 (VEC32_UE8M0, scaleA/B=3) solutions for every K % 64 == 0. Verified on MI355X (gfx950) via hipblaslt-bench: K=64/128/192/256/384 succeed; K=32/96/160/224/288 return "no solution". The 128 bound was over-strict and rejected supported shapes such as K=192. M/N multiple-of-16 requirements are unchanged. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
JohnQinAMD
force-pushed
the
fix/mxfp8-gemm-k-multiple-of-64
branch
from
a435694 to
4092e1e
Compare
ipanfilo
requested changes
Jun 20, 2026
ipanfilo
left a comment
ipanfilo
left a comment
Collaborator
There was a problem hiding this comment.
Corresponding changes in test_cublas_gemm are needed to test k%64 on supported platforms
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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
The ROCm MXFP8 GEMM guard required K % 128 == 0, but hipBLASLt has MXFP8 (VEC32_UE8M0, scaleA/B=3) solutions for every K % 64 == 0. Verified on MI355X (gfx950) via hipblaslt-bench: K=64/128/192/256/384 succeed; K=32/96/160/224/288 return "no solution". The 128 bound was over-strict and rejected supported shapes such as K=192. M/N multiple-of-16 requirements are unchanged.
Description
Please include a brief summary of the changes, relevant motivation and context.
Fixes # (issue)
Type of change
Changes
Please list the changes introduced in this PR:
Checklist: