Разработка программно-математического обеспечения для параллельной обработки разреженных матриц с помощью технологи OpenMP

Abstract

The work is dedicated to parallel computing sparse matrices. The work has been designed and optimized using the C programming language algorithms for working with sparse matrices for the following operations: transposition, multiplication of sparse matrices and sparse matrix multiplication on a tight vector addition of sparse matrices. For data storage it developed a flexible algorithm that uses the most appropriate option to save for a specific type of operation. Selection procedures were applied in dense submatrices sparse matrices. To speed up the program using parallel computing technology OpenMP, which makes it possible to minimize the hardware resources and execution time.