关于DDA算法的并行执行

ON THE PARALLEL EXECUTION OF DDA ALGORITHM

并行处理是使大量处理机尽可能快地执行一个并行程序。如何合理地确定程序的并行粒度是并行处理机系统设计的关键。为了实现DDA 算法的并行处理,本文运用数据流模型表达该算法的计算过程,建立了一种改进的DDA 算法(PDDA),并对其并行性的有效发掘作了研究,提出了一种适宜的粗粒度并行思想,这个思想作为无耦合并行处理系统PACS研究的一部分已经得到验证。

How to determine the granularity of parallelism of a program is a key tothe design of parallel processing systems.For the parallelization of an algorithm——Digital Differential Analyser(DDA),this paper establishes an improved DDAalgorithm(PDDA——Parallel DDA).On the basis of investigation of an efficientexploitation of PDDA's parallelism,an idea of medium granularity of PDDA'sparallelism is presented which,as a portion of design of a decoupled parallelproceesing system PACS,has been verified.