Abstract
SUMMARYThere is an increasing interest in the distribution of parallel random number streams in the high‐performance computing community particularly, with the manycore shift. Even if we have at our disposal statistically sound random number generators according to the latest and thorough testing libraries, their parallelization can still be a delicate problem. Indeed, a set of recent publications shows it still has to be mastered by the scientific community. With the arrival of multi‐core and manycore processor architectures on the scientist desktop, modelers who are non‐specialists in parallelizing stochastic simulations need help and advice in distributing rigorously their experimental plans and replications according to the state of the art in pseudo‐random numbers parallelization techniques. In this paper, we discuss the different partitioning techniques currently in use to provide independent streams with their corresponding software. In addition to the classical approaches in use to parallelize stochastic simulations on regular processors, this paper also presents recent advances in pseudo‐random number generation for general‐purpose graphical processing units. The state of the art given in this paper is written for simulation practitioners. Copyright © 2012 John Wiley & Sons, Ltd.
Highlights
With the arrival of multi-core and manycore processor architectures on the scientist desktop, modelers who are non-specialists in parallelizing stochastic simulations need help and advice in distributing rigorously their experimental plans and replications according to the state of the art in pseudo-random numbers partitioning techniques
For a Pseudo-random number generators (PRNGs) with a period P, the probability that n sequences of length L, generated by a random spacing technique will overlap is equal to 1 − (1 − nL/(P − 1))n−1, which is equivalent to n(n − 1)L/P when nL/P is in the neighborhood of 0
To propose multiple stochastic streams through a general-purpose graphics processing units (GP-GPUs) architecture, we suggest the use of the dynamic creator (DC) dedicated to the new MTGP generator proposed by Saito (as a GP-GPU implementation of Mersenne Twister Saito and Matsumoto (2013))
Summary
With the arrival of multi-core and manycore processor architectures on the scientist desktop, modelers who are non-specialists in parallelizing stochastic simulations need help and advice in distributing rigorously their experimental plans and replications according to the state of the art in pseudo-random numbers partitioning techniques. True random numbers can be clumsy to produce and use, some devices are subject to partial breakdowns (their production has to be regularly statistically checked), some have biases that have to be corrected, and most of the time, we need to store the sequences to be able to reproduce exactly the same sequence several times. This is not possible for many high-performance computing (HPC) applications. In the last 2 sections, we present considerations raised by the emergence of hybrid computing with GP-GPUs
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call