Adding support for bmm_fp8 OP using cutlass#108
Open
bhargaveede wants to merge 3 commits intosgl-project:mainfrom
Open
Adding support for bmm_fp8 OP using cutlass#108bhargaveede wants to merge 3 commits intosgl-project:mainfrom
bhargaveede wants to merge 3 commits intosgl-project:mainfrom
Conversation
msinnha1
reviewed
Feb 24, 2026
| B_scale: torch.Tensor, | ||
| ) -> None: | ||
| cublas_handle = torch.cuda.current_blas_handle() | ||
| # cublas_handle = torch.cuda.current_blas_handle() |
There was a problem hiding this comment.
review comments not resolved from the earlier PR#40 (#40)
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Created new PR inplace of closed PR : #40
@kareemshaik80 @mingfeima This will still be relevant for HW with native support. As we will not do dtype conversions there.
Also, Right now we are doing replication of scales to support stride of 1 (current limitation of cutlass).
On top of current PR, We can avoid that replication once that stride support of 0 is available which will improve the perf further.