Random number generators for massively parallel simulations on GPU

TL;DR

This paper reviews existing GPU-based random number generators and introduces a new high-quality, high-performance generator optimized for massively parallel simulations with minimal memory overhead.

Contribution

It presents a new CUDA-based random number generator designed specifically for highly parallel applications, addressing limitations of existing generators.

Findings

The new generator offers high quality and performance for GPU simulations.

Existing generators are often unsuitable for highly parallel applications due to memory or efficiency issues.

The paper provides a comprehensive review of CUDA random number generator implementations.

Abstract

High-performance streams of (pseudo) random numbers are crucial for the efficient implementation for countless stochastic algorithms, most importantly, Monte Carlo simulations and molecular dynamics simulations with stochastic thermostats. A number of implementations of random number generators has been discussed for GPU platforms before and some generators are even included in the CUDA supporting libraries. Nevertheless, not all of these generators are well suited for highly parallel applications where each thread requires its own generator instance. For this specific situation encountered, for instance, in simulations of lattice models, most of the high-quality generators with large states such as Mersenne twister cannot be used efficiently without substantial changes. We provide a broad review of existing CUDA variants of random-number generators and present the CUDA implementation of a new massively parallel high-quality, high-performance generator with a small memory load overhead.