A novel fault-tolerant routing algorithm in hypercube multi-computers system is proposed,in which each node uses a maximum safety-link matrices(MSLMs) to record the paths to other nodes.

Hypercube multi-computers system is of good performance in parallel and distributed computation.

An Optimized Design of Data Cache in a Multiprocessor System;

A model of system simulation based on FCFS queue dispatching method for multiprocessor system;

In order to develop the performance of the multiprocessor systems, the support software is essential, of which the operating system is the most important.

This paped uses a two-stage arbitration scheme to infer a new performance model for multiprocessor systems employing partial buses as the interconnection networks.

The performance of prioritized nonhomogeneous multiprocessor systems employing (N * N)crossbar as the interconnection networks is studied.

This paper presents a new architecture for multiproces-sor systems,called hierarchical multiple bus multiprocessor,which is applied for the medium and large size of multiprocessor systems.

Elaborated the using of traditional method to enhance the network handling ability and the computation intensity deficiency;analyzed the good and bad points of the multi-processor system as the network server and proposed several methods to overcome its shortcoming

This paper focuses on the designing of debugger for multi-processor system based on four TMS320C25 DSPs, presents the soft-interface method (SIM) and the method of recording breakpoints outside the processor抯 memory (MRBOP).

Cache consistency in the multi-processor system based on centralized management;