Skip to content

Conversation

@simone-silvestri
Copy link
Collaborator

tartarus is quite unused lately. Since we are builing new examples that run on the GPU and for the moment the docs are broken because of an out of memory issue, I thought it might be a good idea to distribute the runs over 2 GPUs.
This might be helpful since we do not always run the examples.

@simone-silvestri simone-silvestri added the build docs Add this label to built the docs in a PR label Dec 23, 2025
JULIA_NUM_PRECOMPILE_TASKS: 8
JULIA_NUM_THREADS: 8
CUDA_VISIBLE_DEVICES: "2" # Device for building ClimaOcean docs
CUDA_VISIBLE_DEVICES: "0,1,2,3" # Devices for building ClimaOcean docs (at the moment only 0, 1 and 2 are used)
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Don't we need 1 or 2 for benchmarking?

Copy link
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

oh right!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

build docs Add this label to built the docs in a PR

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants