Skip to content

Add sparse-sparse transposed MM kernels - #2541

Closed
BarisAygen wants to merge 1 commit into
apache:mainfrom
BarisAygen:sparse-sparse-transposed-mm
Closed

BarisAygen wants to merge 1 commit into
apache:mainfrom
BarisAygen:sparse-sparse-transposed-mm

Conversation

@BarisAygen

Copy link
Copy Markdown
Contributor

Summary

  • Add allocation-free sparse-sparse transposed MM kernels in LibMatrixMult
  • Support A*B, t(A)B, At(B), t(A)*t(B)
  • Add correctness and performance component tests

Support all four transpose combinations without temporary allocation.
@mboehm7

mboehm7 commented Sep 29, 2026

Copy link
Copy Markdown
Contributor

LGTM - thanks for the patch @BarisAygen. During the merge, I moved the performance test to the respective performance benchmarking package (outside our CI).

@mboehm7 mboehm7 closed this in a6d64e8 Sep 29, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

Status: Done

Development

Successfully merging this pull request may close these issues.

2 participants