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.

Sign up for GitHub

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

Jump to bottom

[QST]Why Does CUTLASS Handle the First K Dimension Separately in Matrix Multiplication? #2055

Open
ziyuhuang123 opened this issue Jan 23, 2025 · 1 comment
Open

[QST]Why Does CUTLASS Handle the First K Dimension Separately in Matrix Multiplication? #2055

ziyuhuang123 opened this issue Jan 23, 2025 · 1 comment
Labels
? - Needs Triage question Question

Comments

@ziyuhuang123
Copy link

Could you please explain why CUTLASS separates the first iteration of the K dimension in matrix multiplication? Does this really improve performance?
@hwu36
Copy link
Collaborator

hwu36 commented Feb 6, 2025

For the very least, mma in the prologue can use ScaleOut::Zero to just do C = A x B without adding an existing C accumulator.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
? - Needs Triage question Question
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
2 participants
@hwu36 @ziyuhuang123