Design of the scientific subroutine library for the Fujitsu VPP300

187. R. P. Brent, L. Grosz, D. L. Harrar II, M. Hegland, M. H. Kahn, G. Keating, G. J. Mercer, M. Nakanishi, M. R. Osborne and B. B. Zhou, Design of the scientific subroutine library for the Fujitsu VPP300, Proc. Third High Performance Computing Asia Conference and Exhibition, Singapore, Sept. 1998, 424-438.

Paper: pdf (191K), ps (133K).

Abstract

A research agreement between the Australian National University and Fujitsu Japan has led to the development of a library of parallel mathematical subroutines and the extension of the library of single processor routines for the Fujitsu VPP300. This machine provides a very sophisticated architecture combining vector processors in parallel by a crossbar switch. Very high performance can be obtained if carefully designed algorithms are used which exploit the full capabilities of the hardware.

New algorithms have been developed and performance results are provided here for eigenvalue problems, linear solvers and fast Fourier and wavelet transforms.

Comments

This and the companion paper [182] summarise the main achievements of the "Area 4" Project up to mid-1998. For related work see [132, 141, 148, 153, 155, 163, 170, 185].

Go to next publication

Return to Richard Brent's index page