Remove extra workspace queries in matrix inverse computation #20904
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
Earlier, the workspace size query and allocation was placed inside the loop. However, since we have batches of matrices with the same number of rows and columns, the workspace size query and allocation for every matrix in the batch is redundant. This PR moves the workspace size query and allocation outside the loop, effectively saving (batch_size - 1) number of queries and allocation (and consequently the deallocation). There is a tremendous speedup in inverse computation as a result of this change. Changelog: - Move workspace query and allocation outside the batch loop Test Plan: - All existing tests for inverse should pass to verify that the change is correct
Contributor
Author
|
Speedup summary on batches of 3 x 3 matrices:
|
Contributor
facebook-github-bot
left a comment
facebook-github-bot
left a comment
There was a problem hiding this comment.
@ezyang is landing this pull request. If you are a Facebook employee, you can view this diff on Phabricator.
Contributor
zdevito
pushed a commit
to zdevito/ATen
that referenced
this pull request
May 24, 2019
Summary: Earlier, the workspace size query and allocation was placed inside the loop. However, since we have batches of matrices with the same number of rows and columns, the workspace size query and allocation for every matrix in the batch is redundant. This PR moves the workspace size query and allocation outside the loop, effectively saving (batch_size - 1) number of queries and allocation (and consequently the deallocation). There is a tremendous speedup in inverse computation as a result of this change. Changelog: - Move workspace query and allocation outside the batch loop Pull Request resolved: pytorch/pytorch#20904 Differential Revision: D15495505 Pulled By: ezyang fbshipit-source-id: 226729734465fcaf896f86e1b1a548a81440e082
vishwakftw
added
the
module: performance
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
5 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.
Earlier, the workspace size query and allocation was placed inside the loop.
However, since we have batches of matrices with the same number of rows and columns, the workspace size query and allocation for every matrix in the batch is redundant.
This PR moves the workspace size query and allocation outside the loop, effectively saving (batch_size - 1) number of queries and allocation (and consequently the deallocation).
There is a tremendous speedup in inverse computation as a result of this change.
Changelog:
Test Plan: