Sparse MatrixVector multiplication using parallel algorithm with block striped partitioning

Abstract

Parallel algorithm may be executed as a piece at a time on many processing devices, and then return results at the end. A sparse matrix is huge and consists many zero elements. Its execution process is extremely time-consuming. This paper describes, how to do Sparse Matrix-Vector multiplication using parallel algorithm with block striped partitioning. Speed-up of Sparse Matrix-Vector multiplication measured with increasing number of processors. Experiment results shows that this method increase the speed-up of Sparse Matrix-Vector multiplication