Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Switch to scf::tileUsingSCF #920

Merged
merged 1 commit into from
Jun 5, 2024

Conversation

adam-smnk
Copy link
Collaborator

Deprecates TPP usage of linalg::tileToForallOpUsingTileSizes in preparation for upstream Linalg API deprecation.

The corresponding test is updated as SCF tiling API folds affine maps as part of tiling resulting in simpler IR.

Fixes #676

Deprecates TPP usage of linalg::tileToForallOpUsingTileSizes in
preparation for upstream Linalg API deprecation.

The corresponding test is updated as SCF tiling API folds affine maps
as part of tiling resulting in simpler IR.

Fixes plaidml#676
@adam-smnk adam-smnk requested a review from rengolin June 4, 2024 15:36
@adam-smnk adam-smnk added the benchmark Triggers benchmark jobs label Jun 4, 2024
Copy link
Contributor

@rengolin rengolin left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

That's surprisingly simple! Did the benchmarks look ok?

@adam-smnk
Copy link
Collaborator Author

Did the benchmarks look ok?

Yes, no changes.
Overall, the same IR is generated. The mentioned improvement only occurs with dynamic sizes which we don't really use anywhere apart from a few test cases.

@adam-smnk adam-smnk merged commit 46eff69 into plaidml:main Jun 5, 2024
18 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
benchmark Triggers benchmark jobs
Projects
None yet
Development

Successfully merging this pull request may close these issues.

Canonicalize affine.min map when tensor are dynamics
2 participants