Skip to content

Conversation

@zingale
Copy link
Member

@zingale zingale commented Oct 25, 2025

this gets about 1% performance boost for flame_wave on Frontier

this gets about 1% performance boost for flame_wave on Frontier
@zingale
Copy link
Member Author

zingale commented Oct 25, 2025

curiously this changes answers on groot / CUDA:
http://groot.astro.sunysb.edu/Microphysics/test-suite/gpu/2025-10-25-003/index.html

@zingale
Copy link
Member Author

zingale commented Nov 5, 2025

this seems okay now. We can only put the unrolling into the zero() function without changing the answer on CUDA.

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.

1 participant