The Study On Application Of Parallel Algorithm In The Matrix's Calculation

This paper investigate parallel algorithms and its applications, especially in matrix computing.After introducing the concept of parallel processing and parallel algorithm, we studied the matrix multiplication parallel processing techniques in distributed environment. We analyzed the parallel algorithm for matrix multiplication in MIMD, and described the implementation of the algorithm. We also analyzed the scalability and efficiency of the algorithm.We focused on the study of parallel algorithm for matrix eigenvalue and eigenvector computation. Firstly, we discussed the common used binary approach, then we modified the method and set the scope of eigenvalue for some special cases to simplify the algorithm and to enhance the accelerating ratio. We compared the algorithm with some other ones.We outlooked the algorithm and discussed problems for further studies.