Skip to content

Improve MD performance #6

Description

@lhycms

Motivation

The current MD performance can still be further optimized, especially for GPU inference. Three potential bottlenecks need to be investigated:

  1. The choice of CUDA block_size
  2. The use of atomicAdd for accumulating energy and virial
  3. LAMMPS Kokkos interface

Expected outcome

Reduce GPU overhead in MD inference and improve overall simulation speed, especially for large-scale systems where energy and virial accumulation may become performance bottlenecks.

Metadata

Metadata

Assignees

No one assigned

    Labels

    enhancementNew feature or request

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions