A band relaxation algorithm for reliable and parallelizable circuit simulation

Author

Lumsdaine, A. ; White, J. ; Webber, D. ; Sangiovanni-Vincentelli, A.

Author_Institution

Dept. of Electr. Eng. & Comput. Sci., MIT, Cambridge, MA, USA

fYear

1988

fDate

7-10 Nov. 1988

Firstpage

308

Lastpage

311

Abstract

A variable-band relaxation algorithm for solving large linear systems is developed as an alternative to Gauss-Jacobi relaxation. This algorithm seeks to improve the reliability of Gauss-Jacobi relaxation by extracting a variable-sized band from the matrix and solving that band directly. This leads to a relaxation algorithm with provably better convergence properties. The algorithm can be used effectively on a massively parallel computer because band matrices can be solved in log(n) time on n/2 processors. Test results are presented which compare the convergence properties of variable-band and Gauss-Jacobi relaxation.<>

Keywords

circuit analysis computing; matrix algebra; parallel algorithms; relaxation theory; Gauss-Jacobi relaxation; band matrices; convergence properties; large linear systems; massively parallel computer; parallelizable circuit simulation; variable-band relaxation algorithm; variable-sized band; Circuit simulation; Circuit testing; Computational modeling; Computer science; Convergence; Equations; Gaussian processes; Linear systems; Newton method; SPICE;

fLanguage

English

Publisher

ieee

Conference_Titel

Computer-Aided Design, 1988. ICCAD-88. Digest of Technical Papers., IEEE International Conference on

Conference_Location

Santa Clara, CA, USA

Print_ISBN

0-8186-0869-2

Type

conf

DOI

10.1109/ICCAD.1988.122517

Filename

122517