This paper proposes a fault-tolerant architecture for parallel computer systems, analyzes the fault diagnosis techniques employed in this architecture and presents the configuration optimization strategy for stand-by nodes.

To meet the needs of real time image processing of space applications,a master slave parallel computer system based on MPP techniques is designed in this paper.

With the development of parallel computer systems, the parallel computing systems consisting of a pile of personal computers (PopC for short) have become remarkable recent years due to their high performance- price ratio.

As parallel computers get faster, scientists are increasingly using parallel computers to solve problems where a large amount of computing power is required.

In order to satisfy the strict requirements of application about performance,reliability and cost,the moderate parallel computer system must have high system availability.

Message passing based massive parallel processing systems, which performance greatly depend on the efficiency of the communication among their internal processors, are becoming the mainstream of the supercomputing market.

Design and implementation of a communication processor for parallel computing systems consisting of PC pile;