Overview / Usage

Sparse matrix is generated in almost all large scientific and engineering computing fields, including computational fluid dynamics, statistical physics, circuit simulation, image processing, nanomaterial computing, etc., especially in scientific computing, it is basically a sparse problem. In the field of deep learning, there has also been research on sparse models.

Methodology / Approach

This project is based on Intel oneAPI to realize the parallelization of sparse matrix multiplication in COO format. Through SIMD vectorization, pThread multithreading, and openMP multithreading, we can realize parallelization, greatly improve the speed of sparse matrix multiplication, and compare the advantages and disadvantages of these methods through the comparison of computing time.

Repository

https://github.com/G0atKing/Parallel-Algorithm-of-Sparse-Matrix-Multiplying

Collaborators