A GPU-based large-scale Monte Carlo simulation method for systems with long-range interactions

JOURNAL OF COMPUTATIONAL PHYSICS

卷:338

Article

Liang, Yihao^1,2  Xing, Xiangjun^1,2  Li, Yaohang³ 

[1] Shanghai Jiao Tong Univ, Inst Nat Sci, Shanghai 200240, Peoples R China

[2] Shanghai Jiao Tong Univ, Dept Phys & Astron, Shanghai 200240, Peoples R China

[3] Old Dominion Univ, Dept Comp Sci, Norfolk, VA 23529 USA

关键词: Monte Carlo;    GPU;    Parallel computing;    Coulomb many body systems;    Electrolytes;    Charge renormalization;   

DOI  :  10.1016/j.jcp.2017.02.069

来源: Elsevier

【 摘 要 】

In this work we present an efficient implementation of Canonical Monte Carlo simulation for Coulomb many body systems on graphics processing units (GPU). Our method takes advantage of the GPU Single Instruction, Multiple Data (SIMD) architectures, and adopts the sequential updating scheme of Metropolis algorithm. It makes no approximation in the computation of energy, and reaches a remarkable 440-fold speedup, compared with the serial implementation on CPU. We further use this method to simulate primitive model electrolytes, and measure very precisely all ion-ion pair correlation functions at high concentrations. From these data, we extract the renormalized Debye length, renormalized valences of constituent ions, and renormalized dielectric constants. These results demonstrate unequivocally physics beyond the classical Poisson-Boltzmann theory. (C) 2017 Elsevier Inc. All rights reserved.

【 授权许可】

Free