[cuda] Replace L1 access with warp shuffles in layer_norm_grad_input_kernel #110203
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Outdated suggestions cannot be applied.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.
Replaces accessing L1 for
gamma_val
,c_h
andc_loss
with a warp shuffles. This is guaranteed to work as the unroll factor is lower than the warp width. On average it brings speedups between 0 and 25%.We measured using the benchmark described in #107287
Fixes #ISSUE_NUMBER