[XLA:GPU] Enable wgmma with block_n=8 in Triton and autotuner #97505
Conversation
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
The workaround in AccelerateMatmul forces a contracting-contiguous (K-contiguous) layout for rhs with small N. This prevents from triggering unsupported no-swizzle lowering. We also add checks to WGMMA lowering to catch incorrectly-filled descriptors. Previously this would either crash in runtime or produce incorrect results. PiperOrigin-RevId: 772893153
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.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.
[XLA:GPU] Enable wgmma with block_n=8 in Triton and autotuner
The workaround in AccelerateMatmul forces a contracting-contiguous (K-contiguous) layout for rhs with small N. This prevents from triggering unsupported no-swizzle lowering.
We also add checks to WGMMA lowering to catch incorrectly-filled descriptors. Previously this would either crash in runtime or produce incorrect results.