Skip to content

adds fast blas copy for sparse vectors#51

Merged
dance858 merged 1 commit intomainfrom
fast-path-dense-matmul
Mar 18, 2026
Merged

adds fast blas copy for sparse vectors#51
dance858 merged 1 commit intomainfrom
fast-path-dense-matmul

Conversation

@Transurgeon
Copy link
Copy Markdown
Collaborator

@Transurgeon Transurgeon commented Mar 18, 2026

@dance858 this change seems to effectively bring down the compile time from 50s to around 1s!
really nice intuition that the batched matvec wouldn't have caused such a speedup, and also really nice that we investigated this together.. it helped me understand the code much better :).
Is there a test I should add for this? Any good ideas?

@dance858
Copy link
Copy Markdown
Collaborator

Awesome that such a small change can give such a speedup while keeping the code clean and easy to understand!

@dance858 dance858 merged commit d3a5747 into main Mar 18, 2026
11 checks passed
@Transurgeon Transurgeon deleted the fast-path-dense-matmul branch March 18, 2026 19:23
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants