High-Performance Pseudo-Random Number Generation on Graphics Processing Units
Research School of Computer Science, The Australian National University
arXiv:1108.0486v1 [cs.DC] (2 Aug 2011)
@article{2011arXiv1108.0486N,
author={Nandapalan}, N. and {Brent}, R.~P. and {Murray}, L.~M. and {Rendell}, A.},
title={"{High-Performance Pseudo-Random Number Generation on Graphics Processing Units}"},
journal={ArXiv e-prints},
archivePrefix={"arXiv"},
eprint={1108.0486},
primaryClass={"cs.DC"},
keywords={Computer Science – Distributed, Parallel, and Cluster Computing, Mathematics – Number Theory, Statistics – Computation, 11K45 (Primary) 65C10, 65Y05, 65Y10 (Secondary), D.1.3, G.3, G.4, I.6.8},
year={2011},
month={aug},
adsurl={http://adsabs.harvard.edu/abs/2011arXiv1108.0486N},
adsnote={Provided by the SAO/NASA Astrophysics Data System}
}
This work considers the deployment of pseudo-random number generators (PRNGs) on graphics processing units (GPUs), developing an approach based on the xorgens generator to rapidly produce pseudo-random numbers of high statistical quality. The chosen algorithm has configurable state size and period, making it ideal for tuning to the GPU architecture. We present a comparison of both speed and statistical quality with other common parallel, GPU-based PRNGs, demonstrating favourable performance of the xorgens-based approach.
August 3, 2011 by hgpu
