Skip to content

Conversation

@brownbaerchen
Copy link
Contributor

I migrated everything used for the 3D RBC project so far to its own project directory. This project is now tested on CPUs on Github.
I am relying here on some files from the GPU project. We can either have a self contained project with duplicate code or do this. I opted for the latter. Let me know if you prefer the former.

@pancetta
Copy link
Member

How much do you rely on the GPU project? Does it make sense to move the GPU part to the core routines?

@brownbaerchen
Copy link
Contributor Author

How much do you rely on the GPU project? Does it make sense to move the GPU part to the core routines?

What I am using from there is some infrastructure for running problems i.e. I use commands like mpirun -np 256 python run_experiment --config=RBC... --procs=1/4/64 .... Parts of the GPU project then set up the space time parallelization, add hooks for some logging and output and so on. I think this is specific to my way of executing stuff, so I wouldn't put it in some core place.

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.

2 participants