GPU Optimized Monte Carlo (GOMC) is open-source software for simulating many-body molecular systems using the Metropolis Monte Carlo algorithm. GOMC is written in object oriented C++, which was chosen since it offers a good balance between code development time, interoperability with existing software elements, and code performance. The software may be compiled as a single-threaded application, a multi-threaded application using OpenMP, or to use many-core heterogeneous CPU-GPU architectures using OpenMP and CUDA. GOMC officially supports Windows 7 or newer and most modern distribution of GNU/Linux. This software has the ability to compile on recent versions of macOS; however, such a platform is not officially supported.
GOMC employs widely-used simulation file types (PDB, PSF, CHARMM-style parameter file) and supports polar and non-polar linear and branched molecules. GOMC can be used to study vapor-liquid and liquid-liquid equilibria, adsorption in porous materials, surfactant self-assembly, and condensed phase structure for complex molecules.
GOMC currently is capable of simulating systems in the following ensembles:
- Canonical (NVT)
- Isobaric-isothermal (NPT)
- Grand canonical (μVT)
- Constant volume Gibbs (NVT-Gibbs)
- Constant pressure Gibbs (NPT-Gibbs)
GOMC supports the following Monte Carlo moves: - Rigid-body displacement - Rigid-body rotation - Regrowth using coupled-decoupled configurational-bias - Intra-box swap using coupled-decoupled configurational-bias - Inter-box swap using coupled-decoupled configurational-bias - Volume exchange (both isotropic and anisotropic)
GOMC supports various all-atom, united-atom, and coarse grained force fields: - OPLS - CHARMM - TraPPE - Mie - Martini