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

#18703: Binary_ng - the num of tiles value passed to the LLK CB API's to be a compile-time argument #18645

Open
wants to merge 1 commit into
base: main
Choose a base branch
from

Conversation

KalaivaniMCW
Copy link
Contributor

@KalaivaniMCW KalaivaniMCW commented Mar 5, 2025

Ticket

Link to Github Issue #18703

Problem description

In binary_ng, we are processing onetile per CB for (rd/wr/math). This value can be passed as compile-time arg to the kernels instead of being defined in each kernel. Later if this changes to be a dynamic value, decided by split_work API, we can pass this arg as a runtime-arg to the kernels

What's changed

  • passed num tiles processed by CBs and DST as a compile-time argument
  • move tile_reg api's out of loop to maintain sync between the tile_reg api's

Checklist

@KalaivaniMCW KalaivaniMCW marked this pull request as ready for review March 5, 2025 20:48
@KalaivaniMCW KalaivaniMCW changed the title #0: move tile reg apis out of loop #18703: Binary_ng - the num of tiles value passed to CB LLK API's to be a compile-time argument Mar 6, 2025
@KalaivaniMCW KalaivaniMCW changed the title #18703: Binary_ng - the num of tiles value passed to CB LLK API's to be a compile-time argument #18703: Binary_ng - the num of tiles value passed to the LLK CB API's to be a compile-time argument Mar 6, 2025
@KalaivaniMCW KalaivaniMCW force-pushed the kalaivani/sfpu_kernel_tile branch from 3ec0069 to 5b8eb14 Compare March 7, 2025 15:44
@KalaivaniMCW KalaivaniMCW force-pushed the kalaivani/sfpu_kernel_tile branch from 5b8eb14 to 5ac34f4 Compare March 7, 2025 16:23
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.

4 participants