Add support for smem_epilogue when mma output is not cast to half #3620
+87
−53
Conversation
This file contains 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
|
!test |
jacobhinkle
requested changes
Dec 19, 2024
|
!test |
|
!test |
jacobhinkle
reviewed
Dec 19, 2024
…into pbasu_smem_epi_no_stmatrix
rdspring1
reviewed
Dec 19, 2024
rdspring1
reviewed
Dec 19, 2024
rdspring1
reviewed
Dec 19, 2024
There was a problem hiding this comment.
Refactor Follow-Up Proposal:
It looks like scheduleEpilogue could be broken easily into two functions for readability instead of a single monolithic function.
void HopperMultipleMatmulScheduler::scheduleEpilogueWithVectorization() {}
void HopperMultipleMatmulScheduler::scheduleSmemEpilogue() {}
void HopperMultipleMatmulScheduler::scheduleEpilogue() {
if (!params_->use_smem_epilogue) {
scheduleEpilogueWithVectorization();
} else {
// Use stmatrix (optional) and tma store
scheduleSmemEpilogue();
}
}
|
!test |
|
!test |
|
!test |
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Outdated suggestions cannot be applied.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.
Support non-stmatrix stores from regs to shared memory and then TMA when the output of mma op is not cast back to half precision - stmatrix works with half precision only.