Efficient Sparse Matrix-Matrix Multiplication on Multicore Architectures

Report ID: 
2014-04
Authors: 
Adam Lugowski, John R. Gilbert
Date: 
2014-05-01 00:00:00

Abstract

We describe a new parallel sparse matrix-matrix multiplication algorithm in shared memory using a quadtree decomposition. Our preliminary implementation is nearly as fast as the best sequential method on one core, and scales well to multiple cores.

Document

2014-04.pdf