Regular Article

Massively parallel simulations for disordered systems^★

¹ Institut für Theoretische Physik, Universität Leipzig, IPF 231101, 04081 Leipzig, Germany
² Centre for Fluid and Complex Systems, Coventry University, Coventry CV1 5FB, UK

^a e-mail: martin.weigel@gmail.com

Received: 31 October 2019
Received in final form: 28 February 2020
Published online: 4 May 2020

Abstract

Simulations of systems with quenched disorder are extremely demanding, suffering from the combined effect of slow relaxation and the need of performing the disorder average. As a consequence, new algorithms and improved implementations in combination with alternative and even purpose-built hardware are often instrumental for conducting meaningful studies of such systems. The ensuing demands regarding hardware availability and code complexity are substantial and sometimes prohibitive. We demonstrate how with a moderate coding effort leaving the overall structure of the simulation code unaltered as compared to a CPU implementation, very significant speed-ups can be achieved from a parallel code on GPU by mainly exploiting the trivial parallelism of the disorder samples and the near-trivial parallelism of the parallel tempering replicas. A combination of this massively parallel implementation with a careful choice of the temperature protocol for parallel tempering as well as efficient cluster updates allows us to equilibrate comparatively large systems with moderate computational resources.