This paper discusses program transformations and algorithm modifications that reduce execution time of iterative methods for solv ing partial differential equations on high performance computers. Tech niques typically associated with par allel computers turn out to be essen tial to obtain optimal performance on current superscalar uniprocessors.
Get full access to this article
View all access options for this article.
References
1.
Axelsson, O.1994. Iterative solution methods. Cambridge : Cambridge University Press.
2.
Bagheri, B., Ilin, A., and Scott, L.R.1994. Parallel 3-D MOSFET simulation. In Proc. Hawaii International Conf. SystemSciences, Maui, HI, vol. 1, pp. 46-54.
3.
Bell, R.A.1990. IBM RISC System/6000 performance tuning for numerically intensive FORTRAN and C programs. IBM ITSC Technical Bulletin GG24-3611. IBM United Kingdom Ltd.
4.
Dowd, K.1993. High performance computing. Sebastopol, CA: O'Reilly & Associates.
5.
Ilin, A., Bagheri, B., Scott, L.R., Briggs, J.M., and McCammon, J.A.1995. Parallelization of Poisson-Boltzmann and Brownian dynamics calculation. In Parallel computing in computational chemistry , edited by T. G. Mattson.Washington, DC: ACS Books, pp. 170-185.
6.
Ilin, A., and Scott, L.R.1995a. High performance programming for the KSR-1. Research report UH/MD 196, Department of Mathematics, University of Houston. (Available through http://www.hpc.uh.edu)
7.
Ilin, A., and Scott, L.R.1995b. Loop splitting for superscalar architectures. Research Report UH/MD 195. Department of Mathematics, University of Houston. (Available through http://www.hpc.uh.edu)
8.
Kendall Square Research Corporation.1992. Principles of operation. Waltham, MA: Kendall Square Research.
