A Relational Approach to the Compilation of Sparse Matrix Programs

Abstract. We present a relational algebra based framework for compiling e cient sparse matrix code from dense DO-ANY loops and a speci cation of the representation of the sparse matrix. We present experimental data that demonstrates that the code generated by our compiler achieves performance competitive with that of hand-written codes for important computational kernels.

• Type: Conference
• Year: 1997
• Where: EUROPAR
• Authors: Vladimir Kotlyar, Keshav Pingali, Paul Stodghill
• Keywords: Algebra Based Framework, Distributed And Parallel Computing, EUROPAR 1997, Sparse Matrix, Sparse Matrix Code