The NIST Sparse BLAS (v. 0.9)

Performance Studies on the IBM RS6000 Model 590

Sparse BLAS Homepage





The test matrices used in the following tests were generated by reading sparsity patterns from Harwell-Boeing files, and using these patterns as the block structure for a matrix of given blocksize. The results shown are for Matrix-Vector and Matrix-Matrix multiplications only, rather than a full DAXPY, since we are interested in the efficiency of the sparse code. Source code for the performance testers is available from the authors.

The current test results are for the following matrix patterns: The experimental parameters are blocksize and number of right-hand-sides. We present results of testing on a Sun Sparc 20 and an IBM RS6000. Each data point represents the average result of 4 runs with the same parameters.



Performance of NIST Sparse BLAS on an IBM RS6000:

IMPCOL_C
1 right-hand-side
IBM RS6000 Performance for IMPCOL_C with 1 right-hand-side
10 right-hand-sides
IBM RS6000 Performance for IMPCOL_C with 10 right-hand-sides

WEST0156
1 right-hand-side
IBM RS6000 Performance for WEST0156 with 1 right-hand-side
10 right-hand-sides
IBM RS6000 Performance for WEST0156 with 10 right-hand-sides


GRE 115
1 right-hand-side
IBM RS6000 Performance for GRE 115 with 1 right-hand-side
10 right-hand-sides
IBM RS6000 Performance for GRE 115 with 10 right-hand-sides





Source code for performance testers:




Last updated: July 25, 1996 by KAR.