Skip to content

op - minor at points diagonal improvements#1807

Merged
jeremylt merged 2 commits into
mainfrom
jeremy/extra-ops-trim
Apr 23, 2025
Merged

op - minor at points diagonal improvements#1807
jeremylt merged 2 commits into
mainfrom
jeremy/extra-ops-trim

Conversation

@jeremylt
Copy link
Copy Markdown
Member

Trimming out some extra operations

@jeremylt jeremylt self-assigned this Apr 21, 2025
@jeremylt jeremylt force-pushed the jeremy/extra-ops-trim branch 2 times, most recently from f05046e to 394c0e2 Compare April 21, 2025 21:23
@jeremylt
Copy link
Copy Markdown
Member Author

on my local machine,

$ ./build/ex02-quasistatic -options_file examples/ymls/ex02-quasistatic-elasticity-mpm-neo-hookean-damage-current.yml -use_AT1 true -snes_max_it 5 -expected_strain_energy 1.283335434570e-02 -strain_energy_atol 1e-6 -snes_monitor -ksp_monitor

goes from 2.78ish sec to 1.44ish sec with this change

@jeremylt jeremylt force-pushed the jeremy/extra-ops-trim branch from 394c0e2 to 48fdef1 Compare April 21, 2025 21:29
@jeremylt jeremylt added the GPU label Apr 22, 2025
@jeremylt
Copy link
Copy Markdown
Member Author

Tested the changes for GPU against Ratel too, but I think the kernel launch latency is hiding the real speadups (and a gen version of this will be better anyways)

@jeremylt
Copy link
Copy Markdown
Member Author

Ok, need to find a leak

@jeremylt jeremylt force-pushed the jeremy/extra-ops-trim branch from 8e48a98 to 4101ee3 Compare April 22, 2025 23:35
@jeremylt jeremylt force-pushed the jeremy/extra-ops-trim branch from 4101ee3 to 5cde1db Compare April 23, 2025 14:52
@jeremylt jeremylt merged commit 523ab01 into main Apr 23, 2025
29 checks passed
@jeremylt jeremylt deleted the jeremy/extra-ops-trim branch April 23, 2025 15:56
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant