论文信息 - Unstructured grids on SIMD Torus machines

Unstructured grids on SIMD Torus machines

Unstructured grids lead to unstructured communication on distributed memory parallel computers, a problem that has been considered difficult. We consider adaptive, offline communication routing for a SIMD processor grid. Our approach is empirical. We use large data sets drawn from supercomputing applications instead of an analytic model of communication load. The chief contribution of this paper is an experimental demonstration of the effectiveness of certain routing heuristics. Our routing algorithm is adaptive, nonminimal, and is generally designed to exploit locality. We have a parallel implementation of the router, and we report on its performance.<<ETX>>

P.E. Bjorstad | R. Schreiber

[1] Steven J. Plimpton,et al. An Efficient Parallel Algorithm for Matrix-Vector Multiplication , 1995, Int. J. High Speed Comput..

[2] Sajal K. Das,et al. Book Review: Introduction to Parallel Algorithms and Architectures : Arrays, Trees, Hypercubes by F. T. Leighton (Morgan Kauffman Pub, 1992) , 1992, SIGA.

[3] Joel H. Saltz,et al. Principles of runtime support for parallel processors , 1988, ICS '88.

[4] Steven Warren Hammond,et al. Mapping unstructured grid computations to massively parallel computers , 1992 .

[5] William Aiello,et al. Sparse Matrix Computations on Parallel Processor Arrays , 1993, SIAM J. Sci. Comput..

[6] J. G. Lewis,et al. Distributed memory matrix-vector multiplication and conjugate gradient algorithms , 1993, Supercomputing '93.

[7] Bruce Hendrickson,et al. The Chaco user`s guide. Version 1.0 , 1993 .

[8] E. D. Dahl,et al. Mapping and Compiled Communication on the Connection Machine System , 1990, Proceedings of the Fifth Distributed Memory Computing Conference, 1990..