论文信息 - Template library for multi-GPU pseudorandom number recursion-based generators

Template library for multi-GPU pseudorandom number recursion-based generators

The aim of the paper is to show how to design and implement fast parallel algorithms for Linear Congruential, Lagged Fibonacci and Wichmann-Hill pseudorandom number generators. The new algorithms employ the divide-and-conquer approach for solving linear recurrence systems. They are implemented on multi GPU-accelerated systems using CUDA. Numerical experiments performed on a computer system with two Fermi GPU cards show that our software achieve good performance in comparison to the widely used NVIDIA CURAND Library.

Przemyslaw Stpiczynski | Dominik Szalkowski

[1] Michael Mascagni,et al. SPRNG: A Scalable Library for Pseudorandom Number Generation , 1999, PP.

[2] H. Niederreiter. Quasi-Monte Carlo methods and pseudo-random numbers , 1978 .

[3] Przemyslaw Stpiczynski,et al. Multidimensional Monte Carlo Integration on Clusters with Hybrid GPU-Accelerated Nodes , 2013, PPAM.

[4] Pierre L'Ecuyer,et al. TestU01: A C library for empirical testing of random number generators , 2006, TOMS.

[5] I. D. Hill,et al. Generating good pseudo-random numbers , 2006, Comput. Stat. Data Anal..

[6] Michael Mascagni,et al. Testing parallel random number generators , 2003, Parallel Comput..

[7] J. Mark Bull,et al. Parallel globally adaptive quadrature on the KSR-1 , 1994, Adv. Comput. Math..

[8] Przemyslaw Stpiczynski,et al. Parallel GPU-accelerated recursion-based generators of pseudorandom numbers , 2012, 2012 Federated Conference on Computer Science and Information Systems (FedCSIS).