Add NEON implementation of Fused8BitRowwiseQuantizedSBFloatToFloatOrHalf #3707
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: D69573878 |
✅ Deploy Preview for pytorch-fbgemm-docs ready!
To edit notification comments on pull requests, go to your Netlify site configuration. |
Nicoshev
added a commit
to Nicoshev/FBGEMM
that referenced
this pull request
Feb 18, 2025
…alf (pytorch#3707) Summary: X-link: facebookresearch/FBGEMM#789 QuantUtilsNeon.cc has been introduced, Fused8BitRowwiseQuantizedSBFloatToFloatOrHalfNeon implemented as first function We have observed a ~x12 performance improvement for the downcasting case. The case where a float32_t is returned maintains the same speed: Full results: before: P1732996851 after: P1732996401 Differential Revision: D69573878
Nicoshev
force-pushed
the
export-D69573878
branch
from
99e2504 to
7d12c46
Compare
|
This pull request was exported from Phabricator. Differential Revision: D69573878 |
Nicoshev
added a commit
to Nicoshev/FBGEMM
that referenced
this pull request
Feb 18, 2025
…alf (pytorch#3707) Summary: X-link: facebookresearch/FBGEMM#789 QuantUtilsNeon.cc has been introduced, Fused8BitRowwiseQuantizedSBFloatToFloatOrHalfNeon implemented as first function We have observed a ~x12 performance improvement for the downcasting case. The case where a float32_t is returned maintains the same speed: Full results: before: P1732996851 after: P1732996401 Differential Revision: D69573878
Nicoshev
force-pushed
the
export-D69573878
branch
from
7d12c46 to
3bff7e9
Compare
|
This pull request was exported from Phabricator. Differential Revision: D69573878 |
Nicoshev
added a commit
to Nicoshev/FBGEMM
that referenced
this pull request
Feb 18, 2025
…alf (pytorch#3707) Summary: X-link: facebookresearch/FBGEMM#789 QuantUtilsNeon.cc has been introduced, Fused8BitRowwiseQuantizedSBFloatToFloatOrHalfNeon implemented as first function We have observed a ~x12 performance improvement for the downcasting case. The case where a float32_t is returned maintains the same speed: Full results: before: P1732996851 after: P1732996401 Differential Revision: D69573878
Nicoshev
force-pushed
the
export-D69573878
branch
from
3bff7e9 to
a8931d3
Compare
|
This pull request was exported from Phabricator. Differential Revision: D69573878 |
Nicoshev
force-pushed
the
export-D69573878
branch
from
a8931d3 to
6fdcc64
Compare
Nicoshev
added a commit
to Nicoshev/FBGEMM
that referenced
this pull request
Feb 18, 2025
…alf (pytorch#3707) Summary: X-link: facebookresearch/FBGEMM#789 QuantUtilsNeon.cc has been introduced, Fused8BitRowwiseQuantizedSBFloatToFloatOrHalfNeon implemented as first function We have observed a ~x12 performance improvement for the downcasting case. The case where a float32_t is returned maintains the same speed: Full results: before: P1732996851 after: P1732996401 Differential Revision: D69573878
|
This pull request was exported from Phabricator. Differential Revision: D69573878 |
Nicoshev
added a commit
to Nicoshev/FBGEMM
that referenced
this pull request
Feb 18, 2025
…alf (pytorch#3707) Summary: X-link: facebookresearch/FBGEMM#789 QuantUtilsNeon.cc has been introduced, Fused8BitRowwiseQuantizedSBFloatToFloatOrHalfNeon implemented as first function We have observed a ~x12 performance improvement for the downcasting case. The case where a float32_t is returned maintains the same speed: Full results: before: P1732996851 after: P1732996401 Differential Revision: D69573878
Nicoshev
force-pushed
the
export-D69573878
branch
from
6fdcc64 to
0556e72
Compare
|
This pull request was exported from Phabricator. Differential Revision: D69573878 |
Nicoshev
added a commit
to Nicoshev/FBGEMM
that referenced
this pull request
Feb 18, 2025
…alf (pytorch#3707) Summary: X-link: facebookresearch/FBGEMM#789 QuantUtilsNeon.cc has been introduced, Fused8BitRowwiseQuantizedSBFloatToFloatOrHalfNeon implemented as first function We have observed a ~x12 performance improvement for the downcasting case. The case where a float32_t is returned maintains the same speed: Full results: before: P1732996851 after: P1732996401 Differential Revision: D69573878
Nicoshev
force-pushed
the
export-D69573878
branch
from
0556e72 to
4757d4c
Compare
|
This pull request was exported from Phabricator. Differential Revision: D69573878 |
Nicoshev
added a commit
to Nicoshev/FBGEMM
that referenced
this pull request
Feb 19, 2025
…alf (pytorch#3707) Summary: X-link: facebookresearch/FBGEMM#789 QuantUtilsNeon.cc has been introduced, Fused8BitRowwiseQuantizedSBFloatToFloatOrHalfNeon implemented as first function We have observed a ~x12 performance improvement for the downcasting case. The case where a float32_t is returned maintains the same speed: Full results: before: P1732996851 after: P1732996401 Differential Revision: D69573878
Nicoshev
force-pushed
the
export-D69573878
branch
from
4757d4c to
09852aa
Compare
|
This pull request was exported from Phabricator. Differential Revision: D69573878 |
Nicoshev
added a commit
to Nicoshev/FBGEMM
that referenced
this pull request
Feb 24, 2025
…alf (pytorch#3707) Summary: X-link: facebookresearch/FBGEMM#789 QuantUtilsNeon.cc has been introduced, Fused8BitRowwiseQuantizedSBFloatToFloatOrHalfNeon implemented as first function We have observed a ~x12 performance improvement for the downcasting case. The case where a float32_t is returned maintains the same speed: Full results: before: P1732996851 after: P1732996401 Differential Revision: D69573878
Nicoshev
force-pushed
the
export-D69573878
branch
from
09852aa to
2c5f603
Compare
|
This pull request was exported from Phabricator. Differential Revision: D69573878 |
Nicoshev
added a commit
to Nicoshev/FBGEMM
that referenced
this pull request
Feb 24, 2025
…alf (pytorch#3707) Summary: X-link: facebookresearch/FBGEMM#789 QuantUtilsNeon.cc has been introduced, Fused8BitRowwiseQuantizedSBFloatToFloatOrHalfNeon implemented as first function We have observed a ~x12 performance improvement for the downcasting case. The case where a float32_t is returned maintains the same speed: Full results: before: P1732996851 after: P1732996401 Differential Revision: D69573878
Nicoshev
force-pushed
the
export-D69573878
branch
from
2c5f603 to
2ee6cda
Compare
Nicoshev
added a commit
to Nicoshev/FBGEMM
that referenced
this pull request
Feb 24, 2025
…alf (pytorch#3707) Summary: X-link: facebookresearch/FBGEMM#789 QuantUtilsNeon.cc has been introduced, Fused8BitRowwiseQuantizedSBFloatToFloatOrHalfNeon implemented as first function We have observed a ~x12 performance improvement for the downcasting case. The case where a float32_t is returned maintains the same speed: Full results: before: P1732996851 after: P1732996401 Differential Revision: D69573878
Nicoshev
force-pushed
the
export-D69573878
branch
from
2ee6cda to
cbbf748
Compare
|
This pull request was exported from Phabricator. Differential Revision: D69573878 |
1 similar comment
|
This pull request was exported from Phabricator. Differential Revision: D69573878 |
Nicoshev
added a commit
to Nicoshev/FBGEMM
that referenced
this pull request
Feb 24, 2025
…alf (pytorch#3707) Summary: X-link: facebookresearch/FBGEMM#789 Pull Request resolved: pytorch#3707 QuantUtilsNeon.cc has been introduced, Fused8BitRowwiseQuantizedSBFloatToFloatOrHalfNeon implemented as first function We have observed a ~x12 performance improvement for the downcasting case. The case where a float32_t is returned maintains the same speed: Full results: before: P1732996851 after: P1732996401 Differential Revision: D69573878
Nicoshev
force-pushed
the
export-D69573878
branch
from
cbbf748 to
dfeb094
Compare
…alf (pytorch#3707) Summary: X-link: facebookresearch/FBGEMM#789 QuantUtilsNeon.cc has been introduced, Fused8BitRowwiseQuantizedSBFloatToFloatOrHalfNeon implemented as first function We have observed a ~x12 performance improvement for the downcasting case. The case where a float32_t is returned maintains the same speed: Full results: before: P1732996851 after: P1732996401 Differential Revision: D69573878
Nicoshev
force-pushed
the
export-D69573878
branch
from
dfeb094 to
295e157
Compare
|
This pull request was exported from Phabricator. Differential Revision: D69573878 |
|
This pull request has been merged in 3de6774. |
q10
pushed a commit
to q10/FBGEMM
that referenced
this pull request
Apr 10, 2025
…alf (pytorch#789) Summary: Pull Request resolved: facebookresearch/FBGEMM#789 X-link: pytorch#3707 QuantUtilsNeon.cc has been introduced, Fused8BitRowwiseQuantizedSBFloatToFloatOrHalfNeon implemented as first function We have observed a ~x12 performance improvement for the downcasting case. The case where a float32_t is returned maintains the same speed: Full results: before: P1732996851 after: P1732996401 Reviewed By: q10 Differential Revision: D69573878 fbshipit-source-id: b3d5045cf718a13356c068846a7805ff60f61d87
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:
QuantUtilsNeon.cc has been introduced, Fused8BitRowwiseQuantizedSBFloatToFloatOrHalfNeon implemented as first function
We have observed a ~x12 performance improvement for the downcasting case.
The case where a float32_t is returned maintains the same speed:
Full results:
before:
P1732996851
after:
P1732996401
Differential Revision: D69573878