430786 Development of the Parallel Gibbs Ensemble Monte Carlo Simulation Engine Gomc

Monday, November 9, 2015: 4:25 PM
255A (Salt Palace Convention Center)
Jason R. Mick1, Loren Schwiebert2, Brock Jackman3, Kamel I. Rushaidat2, Mohammad Barhaghi1, Yuanzhe Li2, Younes Nejahi2 and Jeffrey J. Potoff1, (1)Chemical Engineering and Materials Science, Wayne State University, Detroit, MI, (2)Computer Science, Wayne State University, Detroit, MI, (3)Wayne State University, Detroit, MI

GPU Optimized Monte Carlo (GOMC) is a high performance, open source, Gibbs Ensemble Monte Carlo simulation engine that is capable of simulating complex molecular systems containing over 100,000 atoms with performance 10-100 times faster than existing serial codes.  GOMC uses graphics processing units (GPUs) for the calculation of pairwise energies via energetic decomposition.  The code supports simulations in the canonical, isobaric-isothermal, grand canonical and Gibbs ensembles (NVT and NPT) for molecules of arbitrary geometry. GOMC utilizes CHARMM style input files (topology, parameter, PSF, PDB), allowing the use of a single set of input files for CHARMM, NAMD[1] or GOMC.  Output from GOMC is compatible with VMD[2]. 

In this talk, we discuss the implementation of new functionality, which includes support for the calculation of electrostatic energies, new potential functions for non-bonded interactions, and coarse-grained force fields, such as MARTINI.  Optimization strategies unique to simulations with changing numbers of interaction sites, e.g. Gibbs Ensemble and grand canonical Monte Carlo are discussed.   Performance data are presented for serial, OpenMP and GPU versions of the code.

1.         Phillips, J.C., et al., Scalable molecular dynamics with NAMD. Journal of Computational Chemistry, 2005. 26(16): p. 1781-1802.

2.         Humphrey, W., A. Dalke, and K. Schulten, VMD: Visual molecular dynamics. Journal of Molecular Graphics & Modelling, 1996. 14(1): p. 33-38.

Extended Abstract: File Not Uploaded