Suppressing correlations in massively parallel simulations of lattice models
Suppressing correlations in massively parallel simulations of lattice models
Kelling, J.; Ódor, G.; Gemming, S.
For lattice Monte Carlo simulations parallelization is crucial to make studies of large systems and long Simulation time feasible, while sequential simulations remain the gold-standard for correlation-free dynamics. Here, various domain decomposition schemes are compared, concluding with one which delivers virtually correlation-free simulations on GPUs.
Extensive simulations of the octahedron model for 2 + 1 dimensional Kardar–Parisi–Zhang surface growth, which is very sensitive to correlation in the site-selection dynamics, were performed to show self-consistency of the parallel runs and agreement with the sequential algorithm. We present a GPU implementation providing a speedup of about 30× over a parallel CPU implementation on a single socket and at least 180× with respect to the sequential reference.
Keywords: Lattice Monte Carlo; Kardar-Parisi-Zhang; GPU; autocorrelation
-
Computer Physics Communications 220(2017), 205-211
Online First (2017) DOI: 10.1016/j.cpc.2017.07.010
Cited 3 times in Scopus -
Contribution to WWW
https://arxiv.org/abs/1705.01022 -
Software in external data repository
Publication year 2017
Programming language: C++
System requirements: Linux, GCC, CUDA
License: GPLv3+ (Link to license text)
Hosted on GitHub: Link to location
Downloads
- Final Draft PDF 1,1 MB Secondary publication
Permalink: https://www.hzdr.de/publications/Publ-25479