Enable faster cuBLAS path for torch.linalg.lstsq for batch of small matrices#74434
Closed
IvanYashchuk wants to merge 2 commits intopytorch:masterfrom
Closed
Enable faster cuBLAS path for torch.linalg.lstsq for batch of small matrices#74434IvanYashchuk wants to merge 2 commits intopytorch:masterfrom
IvanYashchuk wants to merge 2 commits intopytorch:masterfrom
Conversation
CI Flow Status⚛️ CI FlowRuleset - Version:
|
Contributor
🔗 Helpful links
💊 CI failures summary and remediationsAs of commit 42b868c (more details on the Dr. CI page): 💚 💚 Looks good so far! There are no failures yet. 💚 💚 This comment was automatically generated by Dr. CI (expand for details).Please report bugs/suggestions to the (internal) Dr. CI Users group. |
mruberry
reviewed
Mar 22, 2022
| #endif // AT_MAGMA_ENABLED() | ||
| // On ROCm platform we can only use MAGMA here | ||
| // If MAGMA is not available, an error will be thrown | ||
| gels_magma(a, b, infos); |
Collaborator
There was a problem hiding this comment.
Why change the behavior on ROCm?
Collaborator
Author
There was a problem hiding this comment.
It doesn't change it. It makes it more explicit that on ROCm magma function is called. Previously the chain of calls was "this_function() -> gels_looped() -> gels_magma()", now it's "this_function() -> gels_magma()".
Collaborator
There was a problem hiding this comment.
Cool, thanks for the explanation
mruberry
approved these changes
Mar 22, 2022
Collaborator
mruberry
left a comment
There was a problem hiding this comment.
Cool! -- One question for you, @IvanYashchuk
Collaborator
|
@pytorchbot merge this please |
Contributor
|
Hey @IvanYashchuk. |
facebook-github-bot
pushed a commit
that referenced
this pull request
Mar 23, 2022
…atrices (#74434) Summary: This PR enables cuBLAS path for `torch.linalg.lstsq`. Before this PR only cuSOLVER path was used for regular PyTorch builds (when built with MAGMA). Performance results (also previously reported at #54725 (comment)): ``` | | before current PR | current PR | speedup | |----------------------------|-------------------|------------|---------| | torch.Size([32, 32, 32]) | 870 | 440 | 2x | | torch.Size([64, 32, 32]) | 1340 | 450 | 3x | | torch.Size([32, 64, 64]) | 9040 | 1839 | 5x | | torch.Size([64, 64, 64]) | 17000 | 1830 | 9.2x | | torch.Size([32, 128, 128]) | 23210 | 8560 | 2.7x | | torch.Size([64, 128, 128]) | 40000 | 8662 | 4.6x | | torch.Size([32, 256, 256]) | 58160 | 46150 | 1.2x | | torch.Size([64, 256, 256]) | 73220 | 52080 | 1.4x | Times are in microseconds (us). ``` Pull Request resolved: #74434 Approved by: https://github.com/mruberry Test Plan: contbuild & OSS CI, see https://hud.pytorch.org/commit/pytorch/pytorch/4de870f6040d2799acc11b9bdeb6508eb6fd33d9 Reviewed By: malfet Differential Revision: D35047957 fbshipit-source-id: c21c3fdabcf2fc747089e5915fe66561760602c3
shahofblah
pushed a commit
that referenced
this pull request
Mar 25, 2022
…atrices This PR enables cuBLAS path for `torch.linalg.lstsq`. Before this PR only cuSOLVER path was used for regular PyTorch builds (when built with MAGMA). Performance results (also previously reported at #54725 (comment)): ``` | | before current PR | current PR | speedup | |----------------------------|-------------------|------------|---------| | torch.Size([32, 32, 32]) | 870 | 440 | 2x | | torch.Size([64, 32, 32]) | 1340 | 450 | 3x | | torch.Size([32, 64, 64]) | 9040 | 1839 | 5x | | torch.Size([64, 64, 64]) | 17000 | 1830 | 9.2x | | torch.Size([32, 128, 128]) | 23210 | 8560 | 2.7x | | torch.Size([64, 128, 128]) | 40000 | 8662 | 4.6x | | torch.Size([32, 256, 256]) | 58160 | 46150 | 1.2x | | torch.Size([64, 256, 256]) | 73220 | 52080 | 1.4x | Times are in microseconds (us). ``` Pull Request resolved: #74434 Approved by: https://github.com/mruberry
laurentdupin
pushed a commit
to laurentdupin/pytorch
that referenced
this pull request
Apr 25, 2026
…atrices This PR enables cuBLAS path for `torch.linalg.lstsq`. Before this PR only cuSOLVER path was used for regular PyTorch builds (when built with MAGMA). Performance results (also previously reported at pytorch#54725 (comment)): ``` | | before current PR | current PR | speedup | |----------------------------|-------------------|------------|---------| | torch.Size([32, 32, 32]) | 870 | 440 | 2x | | torch.Size([64, 32, 32]) | 1340 | 450 | 3x | | torch.Size([32, 64, 64]) | 9040 | 1839 | 5x | | torch.Size([64, 64, 64]) | 17000 | 1830 | 9.2x | | torch.Size([32, 128, 128]) | 23210 | 8560 | 2.7x | | torch.Size([64, 128, 128]) | 40000 | 8662 | 4.6x | | torch.Size([32, 256, 256]) | 58160 | 46150 | 1.2x | | torch.Size([64, 256, 256]) | 73220 | 52080 | 1.4x | Times are in microseconds (us). ``` Pull Request resolved: pytorch#74434 Approved by: https://github.com/mruberry
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 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.
This PR enables cuBLAS path for
torch.linalg.lstsq. Before this PR only cuSOLVER path was used for regular PyTorch builds (when built with MAGMA).Performance results (also previously reported at #54725 (comment)):