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

Option to increase problem size of nonlinear solve benchmark (small or large) #1251

Merged
merged 2 commits into from
Oct 25, 2024

Conversation

chapman39
Copy link
Contributor

@chapman39 chapman39 commented Oct 25, 2024

This is a part of improving performance analysis #1226.

This PR...

Adds a -s option to the solid nonlinear solve benchmark, where you can specify the problem size of the benchmark. A SMALL problem size takes under 5 seconds. This could be used to test whether the benchmark is working as intended. A LARGE problem size takes about 5 minutes with 16 tasks. This should be used to measure performance.

The motivation behind increasing the problem size is we want our benchmarks to be large enough to give us more accurate timing information. I figured I would do this before I start adding benchmarks to CI.

I tried adding an extra 0 to the large problem size and got std::bad_alloc:

Nx = 10000;
Ny = 60;
Nz = 50;

@chapman39 chapman39 self-assigned this Oct 25, 2024
@chapman39 chapman39 added the testing Related to testing label Oct 25, 2024
@chapman39 chapman39 marked this pull request as ready for review October 25, 2024 00:43
@white238
Copy link
Member

/style

@chapman39 chapman39 merged commit e229a18 into develop Oct 25, 2024
2 checks passed
@chapman39 chapman39 deleted the feature/chapman39/increase-benchmark-scale branch October 25, 2024 18:20
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
testing Related to testing
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants