arXiv (Cornell University) · 2010 · 11 citations · 11 references
ComputingComputer EngineeringComputer ScienceComputer ArchitectureComputer GraphicsParallel ComputingData ParallelismGraphics Processing Unit ArchitectureGraphics Processing Unit ComputingGraphics Processing UnitsParallel ArchitecturesRandom Number GenerationGpu ArchitectureMersenne TwisterParallel Random Number GenerationPseudo-random Number GeneratorsPseudorandom FunctionsPseudorandom GeneratorsPseudorandom PermutationsRecursion
This paper proposes a type of pseudorandom number generator, Mersenne Twister for Graphic Processor (MTGP), for efficient generation on graphic processessing units (GPUs). MTGP supports large state sizes such as 11213 bits, and uses the high parallelism of GPUs in computing many steps of the recursion in parallel. The second proposal is a parameter-set generator for MTGP, named MTGP Dynamic Creator (MTGPDC). MT- GPDC creates up to 2^32 distinct parameter sets which generate sequences with high-dimensional uniformity. This facility is suitable for a large grid of GPUs where each GPU requires separate random number streams. MTGP is based on linear recursion over the two-element field, and has better high-dimensional equidistribution than the Mersenne Twister pseudorandom number generator.
11
Makoto Matsumoto, Takuji Nishimura · ACM Transactions on Modeling and Computer Simulation · 1998 · 5.7K citations
Pierre L’Ecuyer, Richard Simard · ACM Transactions on Mathematical Software · 2007 · 998 citations
David Luebke, Mark Harris, Jens Krüger et al. · 2004 · 228 citations