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

Gather/scatter2 #43

Merged
merged 9 commits into from
Jun 23, 2024
Merged

Gather/scatter2 #43

merged 9 commits into from
Jun 23, 2024

Conversation

Joroks
Copy link
Collaborator

@Joroks Joroks commented Jun 22, 2024

adresses #41. @vchuravy

I've started from scratch with my gather/scatter implementation, because I've overlooked some part of the C-API perviously that allows me to determine the count regardless of what type of data gets gathered/scattered. I therefore also removed count from the parameter list.

src/LAMMPS.jl Outdated Show resolved Hide resolved
src/LAMMPS.jl Outdated Show resolved Hide resolved
@vchuravy vchuravy merged commit 212af20 into cesmix-mit:main Jun 23, 2024
8 of 15 checks passed
@vchuravy
Copy link
Member

Thanks! This is a great contribution.

I was wondering what your plans are with LAMMPS.jl
Would something like #28 be helpful in your work?

@Joroks
Copy link
Collaborator Author

Joroks commented Jun 23, 2024

You're welcome! I think gather and scatter is basically everything that I need for my project at the moment. I'm currently working on Implementing a Finite-Element-Molecular-Dynamics coupling method, that alternates between FE and MD simulation steps. So after each iteration (either FE or MD) I have to update the positions and forces on certain atoms designated as anchors.

The original implementation of that Method (written in Matlab) used to read/write this information from/to dump files while starting a new LAMMPS instance for each MD iteration step. With this change I don't have to bother with reading/writing files while also hopefully removing the overhead from restarting LAMMPS for each Iteration Step.

I'm hopeing that this will give me a nice performance boost - apart from the ~100x performance increase I was already able to get from writing the FE simulation in Julia (using Ferrite) instead of Matlab.

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