Optimize tbe_input_combine_with_length_cuda on AMD #4430
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
✅ Deploy Preview for pytorch-fbgemm-docs ready!
To edit notification comments on pull requests, go to your Netlify project configuration. |
|
This pull request was exported from Phabricator. Differential Revision: D75886673 |
JChunX
force-pushed
the
export-D75886673
branch
2 times, most recently
from
3cd28ce to
0f2c26c
Compare
|
This pull request was exported from Phabricator. Differential Revision: D75886673 |
Summary: Pull Request resolved: pytorch#4430 X-link: facebookresearch/FBGEMM#1496 DPER frontend benchmark show tbe_input_combine_with_length_cuda as one of the top contributors to local net latency on CMF fully remote model. Especially on AMD, where latency for this kernel is ~2x of NVIDIA (albeit AMD executes with more kernels in parallel). VEC_WIDTH=32 Increase items processed on AMD per thread, improving memory access patterns and taking advantage of AMD GPU larger memory bandwidth. Reviewed By: q10 Differential Revision: D75886673
|
This pull request was exported from Phabricator. Differential Revision: D75886673 |
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Outdated suggestions cannot be applied.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.
Summary:
DPER frontend benchmark show tbe_input_combine_with_length_cuda as one of the top contributors to local net latency on CMF fully remote model. Especially on AMD, where latency for this kernel is ~2x of NVIDIA (albeit AMD executes with more kernels in parallel).
VEC_WIDTH=32 Increase items processed on AMD per thread, improving memory access patterns and taking advantage of AMD GPU larger memory bandwidth.
Differential Revision: D75886673