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

optimize GPU force; match GPU force w/CPU force in all cases #40

Open
khavernathy opened this issue Nov 14, 2017 · 1 comment
Open

optimize GPU force; match GPU force w/CPU force in all cases #40

khavernathy opened this issue Nov 14, 2017 · 1 comment

Comments

@khavernathy
Copy link
Owner

khavernathy commented Nov 14, 2017

This needs a thorough check.

Consolidate pair distance calculator to 1x instead of 3x for polarized force.
Likewise consolidate local (accumulated) force vector to save time by avoiding atomicAdd()
LJ should (probably) check mixing before computing r.

@khavernathy khavernathy changed the title match GPU force w/CPU force in all cases optimize GPU force; match GPU force w/CPU force in all cases Nov 16, 2017
@khavernathy
Copy link
Owner Author

The (small) discrepancy between GPU and CPU force emerges from NVIDIA CUDA compiler forcing floats instead of doubles.

Need to optimize/clean CUDA functions for force calculation

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

1 participant