Use cudaMemset/hipMemset to setup IndexShuffling kernel. #4016
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
|
This pull request was exported from Phabricator. Differential Revision: D73602755 |
✅ Deploy Preview for pytorch-fbgemm-docs ready!
To edit notification comments on pull requests, go to your Netlify site configuration. |
levendlee
force-pushed
the
export-D73602755
branch
from
c2bbc70 to
282cf97
Compare
levendlee
added a commit
to levendlee/FBGEMM
that referenced
this pull request
Apr 24, 2025
Summary: X-link: facebookresearch/FBGEMM#1104 It is too expensive to launch a ATen kernel to do setup. Use cudaMemset/hipMemset instead. Differential Revision: D73602755
levendlee
added a commit
to levendlee/FBGEMM
that referenced
this pull request
Apr 24, 2025
Summary: X-link: facebookresearch/FBGEMM#1104 It is too expensive to launch a ATen kernel to do setup. Use cudaMemset/hipMemset instead. Differential Revision: D73602755
levendlee
force-pushed
the
export-D73602755
branch
from
282cf97 to
6f57ffb
Compare
levendlee
added a commit
to levendlee/FBGEMM
that referenced
this pull request
Apr 24, 2025
Summary: X-link: facebookresearch/FBGEMM#1104 It is too expensive to launch a ATen kernel to do setup. Use cudaMemset/hipMemset instead. Differential Revision: D73602755
levendlee
force-pushed
the
export-D73602755
branch
from
6f57ffb to
fce075c
Compare
|
This pull request was exported from Phabricator. Differential Revision: D73602755 |
1 similar comment
|
This pull request was exported from Phabricator. Differential Revision: D73602755 |
levendlee
force-pushed
the
export-D73602755
branch
from
fce075c to
807fd87
Compare
levendlee
added a commit
to levendlee/FBGEMM
that referenced
this pull request
Apr 24, 2025
Summary: Pull Request resolved: pytorch#4016 X-link: facebookresearch/FBGEMM#1104 It is too expensive to launch a ATen kernel to do setup. Use cudaMemset/hipMemset instead. Differential Revision: D73602755
|
This pull request was exported from Phabricator. Differential Revision: D73602755 |
levendlee
added a commit
to levendlee/FBGEMM
that referenced
this pull request
Apr 24, 2025
Summary: Pull Request resolved: pytorch#4016 X-link: facebookresearch/FBGEMM#1104 It is too expensive to launch a ATen kernel to do setup. Use cudaMemset/hipMemset instead. Differential Revision: D73602755
levendlee
force-pushed
the
export-D73602755
branch
2 times, most recently
from
a783b2b to
ab6f083
Compare
levendlee
added a commit
to levendlee/FBGEMM
that referenced
this pull request
Apr 24, 2025
Summary: X-link: facebookresearch/FBGEMM#1104 It is too expensive to launch a ATen kernel to do setup. Use cudaMemset/hipMemset instead. Differential Revision: D73602755
levendlee
force-pushed
the
export-D73602755
branch
from
ab6f083 to
21b0fb0
Compare
levendlee
added a commit
to levendlee/FBGEMM
that referenced
this pull request
Apr 24, 2025
Summary: X-link: facebookresearch/FBGEMM#1104 It is too expensive to launch a ATen kernel to do setup. Use cudaMemset/hipMemset instead. Differential Revision: D73602755
|
This pull request was exported from Phabricator. Differential Revision: D73602755 |
levendlee
added a commit
to levendlee/FBGEMM
that referenced
this pull request
Apr 24, 2025
Summary: Pull Request resolved: pytorch#4016 X-link: facebookresearch/FBGEMM#1104 It is too expensive to launch a ATen kernel to do setup. Use cudaMemset/hipMemset instead. Differential Revision: D73602755
levendlee
force-pushed
the
export-D73602755
branch
from
21b0fb0 to
cf1e7e2
Compare
|
This pull request was exported from Phabricator. Differential Revision: D73602755 |
levendlee
added a commit
to levendlee/FBGEMM
that referenced
this pull request
Apr 24, 2025
Summary: Pull Request resolved: pytorch#4016 X-link: facebookresearch/FBGEMM#1104 It is too expensive to launch a ATen kernel to do setup. Use cudaMemset/hipMemset instead. Differential Revision: D73602755
levendlee
force-pushed
the
export-D73602755
branch
from
cf1e7e2 to
f630cde
Compare
levendlee
added a commit
to levendlee/FBGEMM
that referenced
this pull request
Apr 25, 2025
Summary: X-link: facebookresearch/FBGEMM#1104 It is too expensive to launch a ATen kernel to do setup. Use cudaMemset/hipMemset instead. Differential Revision: D73602755
levendlee
force-pushed
the
export-D73602755
branch
from
f630cde to
02cb1da
Compare
levendlee
added a commit
to levendlee/FBGEMM
that referenced
this pull request
Apr 25, 2025
Summary: X-link: facebookresearch/FBGEMM#1104 It is too expensive to launch a ATen kernel to do setup. Use cudaMemset/hipMemset instead. Differential Revision: D73602755
levendlee
force-pushed
the
export-D73602755
branch
from
5a56152 to
513eb3a
Compare
levendlee
added a commit
to levendlee/FBGEMM
that referenced
this pull request
Apr 25, 2025
…ytorch#4016) Summary: X-link: facebookresearch/FBGEMM#1104 It is too expensive to launch a ATen kernel to do setup. Use cudaMemsetAsync/hipMemsetAsync instead. Note we need to use cudaMemsetAsync/hipMemsetAsync on current stream to be compatible with CUDAGraph capture. Reviewed By: Alkaid-Benetnash Differential Revision: D73602755
levendlee
force-pushed
the
export-D73602755
branch
from
513eb3a to
2b1606e
Compare
|
This pull request was exported from Phabricator. Differential Revision: D73602755 |
1 similar comment
|
This pull request was exported from Phabricator. Differential Revision: D73602755 |
levendlee
added a commit
to levendlee/FBGEMM
that referenced
this pull request
Apr 25, 2025
…ytorch#4016) Summary: Pull Request resolved: pytorch#4016 X-link: facebookresearch/FBGEMM#1104 It is too expensive to launch a ATen kernel to do setup. Use cudaMemsetAsync/hipMemsetAsync instead. Note we need to use cudaMemsetAsync/hipMemsetAsync on current stream to be compatible with CUDAGraph capture. Reviewed By: Alkaid-Benetnash Differential Revision: D73602755
levendlee
force-pushed
the
export-D73602755
branch
from
2b1606e to
8cb941b
Compare
|
This pull request was exported from Phabricator. Differential Revision: D73602755 |
levendlee
force-pushed
the
export-D73602755
branch
from
8cb941b to
d6a9b6f
Compare
levendlee
added a commit
to levendlee/FBGEMM
that referenced
this pull request
Apr 25, 2025
…ytorch#4016) Summary: Pull Request resolved: pytorch#4016 X-link: facebookresearch/FBGEMM#1104 It is too expensive to launch a ATen kernel to do setup. Use cudaMemsetAsync/hipMemsetAsync instead. Note we need to use cudaMemsetAsync/hipMemsetAsync on current stream to be compatible with CUDAGraph capture. Reviewed By: Alkaid-Benetnash Differential Revision: D73602755
levendlee
force-pushed
the
export-D73602755
branch
from
d6a9b6f to
046f3f4
Compare
levendlee
added a commit
to levendlee/FBGEMM
that referenced
this pull request
Apr 27, 2025
Summary: X-link: facebookresearch/FBGEMM#1104 It is too expensive to launch a ATen kernel to do setup. Use cudaMemsetAsync instead. hipMemsetAsync is somehow more expensive than launching a kernel. Avoid doing so for now. Reviewed By: Alkaid-Benetnash Differential Revision: D73602755
|
This pull request was exported from Phabricator. Differential Revision: D73602755 |
levendlee
added a commit
to levendlee/FBGEMM
that referenced
this pull request
Apr 28, 2025
Summary: X-link: facebookresearch/FBGEMM#1104 It is too expensive to launch a ATen kernel to do setup. Use cudaMemsetAsync instead. hipMemsetAsync is somehow more expensive than launching a kernel. Avoid doing so for now. Reviewed By: Alkaid-Benetnash Differential Revision: D73602755
levendlee
force-pushed
the
export-D73602755
branch
from
046f3f4 to
f36a72e
Compare
levendlee
added a commit
to levendlee/FBGEMM
that referenced
this pull request
Apr 28, 2025
Summary: X-link: facebookresearch/FBGEMM#1104 It is too expensive to launch a ATen kernel to do setup. Use cudaMemsetAsync instead. hipMemsetAsync is somehow more expensive than launching a kernel. Avoid doing so for now. Reviewed By: Alkaid-Benetnash Differential Revision: D73602755
levendlee
force-pushed
the
export-D73602755
branch
from
f36a72e to
1d6e112
Compare
levendlee
added a commit
to levendlee/FBGEMM
that referenced
this pull request
Apr 28, 2025
Summary: X-link: facebookresearch/FBGEMM#1104 It is too expensive to launch a ATen kernel to do setup. Use cudaMemsetAsync instead. hipMemsetAsync is somehow more expensive than launching a kernel. Avoid doing so for now. Reviewed By: Alkaid-Benetnash Differential Revision: D73602755
levendlee
force-pushed
the
export-D73602755
branch
from
1d6e112 to
0cb545f
Compare
|
This pull request was exported from Phabricator. Differential Revision: D73602755 |
1 similar comment
|
This pull request was exported from Phabricator. Differential Revision: D73602755 |
levendlee
force-pushed
the
export-D73602755
branch
from
0cb545f to
ade50d9
Compare
levendlee
added a commit
to levendlee/FBGEMM
that referenced
this pull request
Apr 28, 2025
Summary: Pull Request resolved: pytorch#4016 X-link: facebookresearch/FBGEMM#1104 It is too expensive to launch a ATen kernel to do setup. Use cudaMemsetAsync instead. hipMemsetAsync is somehow more expensive than launching a kernel. Avoid doing so for now. Reviewed By: Alkaid-Benetnash Differential Revision: D73602755
Summary: Pull Request resolved: pytorch#4016 X-link: facebookresearch/FBGEMM#1104 It is too expensive to launch a ATen kernel to do setup. Use cudaMemsetAsync instead. hipMemsetAsync is somehow more expensive than launching a kernel. Avoid doing so for now. Reviewed By: Alkaid-Benetnash Differential Revision: D73602755
|
This pull request was exported from Phabricator. Differential Revision: D73602755 |
levendlee
force-pushed
the
export-D73602755
branch
from
ade50d9 to
a331692
Compare
|
This pull request has been merged in eeee38e. |
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:
X-link: https://github.com/facebookresearch/FBGEMM/pull/1104
It is too expensive to launch a ATen kernel to do setup. Use cudaMemset/hipMemset instead.
Differential Revision: D73602755