The emergence of multi-core technology proposes a method for people to speed the computer. A large number of parallel algorithms are designed and applied to many situations as well. It aims to design a new parallel algorithm which used in CRNG (Combined Random Number Generator) to improve the computing speed of the traditional method. In this paper,the traditional combination of random number generator computing task decomposition process which based on the parallel model in the level of task, its execution cores assigned to four parallel execution,in order to produce the final sequence of random numbers. Finally,use the common tool of parallel programming-OpenMP programming for the new algorithm validation in the Windows environment,the results confirmed that the algorithm can make full use of existing computer resources can provide multi-core computing and accelerate more than 3 times.%多核技术的出现给人们带来了一种大幅提升计算机运行速度的方法,大量的并行算法也被设计并应用到各个场合中.文中目的在于设计一种新的用于组合随机数发生器CRNG( Combined Random Number Generator)的并行算法,以提高传统算法的运算速率.文中采用并行编程方法中的任务级的并行模式,对传统组合随机数发生器的运算过程进行任务分解,将其分配到四个执行核上并行执行,以产生最终的随机数序列.最后在Windows环境下,使用常用的并行编程工具-OpenMP对新算法进行了编程验证,结果证实该算法可充分利用现有计算机所能提供的多核计算资源,其加速比高于3.
展开▼
