PMEOPA, a Parallel Molecular Energetics using Octree-based Pairwise Approximation, have been implemented in order to perform fast computation of molecular mechanical (Lennard-Jones and Coulombic) and generalized Born solvation energetics which run on commodity multi-core CPUs and GPUs.
- Supported computations
- Lennard-Jones, Coulombic, molecular surface based effective Born radii, generalized Born polarization
- For each terms,
- parallel computation on GPUs using CUDA
- parallel computation on multi-core using Cilk++
- naive (n^2) computation for a comparison
R. Chowdhury and C. Bajaj
Algorithms for faster molecular energetics, forces and interfaces
ICES Technical Report 10-32, The University of Texas at Austin, 2010.
- Two main dependencies of PMEOPA are Nvidia CUDA (https://developer.nvidia.com/cuda-downloads) and Intel Cilk++ (http://software.intel.com/en-us/articles/download-intel-cilk-sdk)
- Here is the user guide for PMEOPA which includes detailed installation and usage guide.