Research And Achieve Of The Parallel Algorithm Of The Matrix Computing On More Processors

High performance computer technology is comprehensive reflection of a national scientific technology and economic strength, and is an important tool to promoting the advancement of scientific technology and economic strength, national security and various work of society. With the more progress of high-performance computer technology, the higher requirementsof computer about computational accuracy, processing speed and many performance are put forward, so computer cluster system based on distributed storage technology and cluster technology will be slowly become a effective method of meeting the high-performance data processing requirements.matrix calculation operations are used in many Numerical computing operations, so that the matrix operations is the one of important numerical computing operations, especially for high matrix operation, its calculation takes more processing units and larger memory unit, so the efficiency of numerical calculation will be greatly affected. Therefore, the research of parallel of matrix operations is very significant.This article firstly introduces the parallel computer system, the the building of computer cluster environment, parallel algorithm, the design of parallel program and so on, and then introduces the basic knowledgy of MPI programming and the establishment of minicomputer cluster environment based on Windows platform, followed introduces about the parallel algorithm of matrix transpose, matrix-vector multiplication, matrix inverse finally, this paper sums up the lack of the work and the problems of needing to continue research.