Abstract
Given a n×n nonsingular linear system Ax=b, we prove that the solution x can be computed in parallel time ranging from Ω(log n) to O(log2 n), provided that the condition number, μ(A), of A is bounded by a polynomial in n. In particular, if μ(A) = O(1), a time bound O(log n) is achieved. To obtain this result, we reduce the computation of x to repeated matrix squaring and prove that a number of steps independent of n is sufficient to approximate x up to a relative error 2−d, d=O(1). This algorithm has both theoretical and practical interest, achieving the same bound of previously published parallel solvers, but being far more simple.
This work has been partly supported by the Italian National Research Council, under the “Progetto Finalizzato Sistemi Informatici e Calcolo Parallelo”, subproject 2 “Processori dedicati”. Part of this work was done while the first author was with the Istituto di Elaborazione dell'Informazione, Consiglio Nazionale delle Ricerche, Pisa (Italy).
This is a preview of subscription content, log in via an institution.
Preview
Unable to display preview. Download preview PDF.
References
S. J. Berkowitz. On computing the determinant in small parallel time using a small number of processors. Inform. Process. Lett., 18:147–150, 1984.
B. Codenotti. Parallel solution of linear systems by repeated squaring. Applied Math. Letters, 3:19–20, 1990.
B. Codenotti and F. Flandoli. A monte carlo method for the parallel solution of linear systems. Journal of Complexity, 5:107–117, 1989.
B. Codenotti and M. Leoncini. Parallel Complexity of Linear System Solution. World Scientific Pu. Co., Singapore, 1991.
S. A. Cook. A taxonomy of problems with fast parallel algorithms. Inform. and Control, 64:2–22, 1985.
D. Coppersmith and S. Winograd. Matrix multiplication via arithmetic progression. In Proc. 19th Annual ACM Symposium on Theory of Computing, pages 1–6, Berkeley, CA, 1987. Springer-Verlag.
M. Cosnard and Y. Robert. Complexity of parallel QR factorization. J. Assoc. Cornput. Mach., 33:712–723, 1986.
L. Csanky. Fast parallel matrix inversion algorithms. SIAM J. Comput., 5:618–623, 1976.
J. von zur Gathen. Parallel arithmetic computations: a survey. In Lecture notes in Computer Science, volume 233, pages 93–122. Springer-Verlag, New-York, 1986.
J. von zur Gathen. Parallel linear algebra. In J. Reif, editor, Synthesis of Parallel Algorithms. 1991. to appear.
G. H. Golub and C. F. van Loan. Matrix Computations. Johns Hopkins University Press, Baltimore, Md, 1989.
V. Pan. Complexity of parallel matrix computations. Theoret. Comput. Sci., 54:65–85, 1987.
V. Pan and J. Reif. Efficient parallel solution of linear systems. In Proc. 17th Annual ACM Symposium on Theory of Computing, pages 143–152, 1985.
V. Pan and J. Reif. Fast and efficient parallel solution of dense linear systems. Comput. and Math. with Appl., 17:1481–1491, 1989.
F. P. Preparata. Inverting a Vandermonde matrix in minimum parallel time. Inform. Process. Lett., 38:291–294, 1991.
A. H. Sameh and D. J. Kuck. On stable parallel linear systems solvers. J. Assoc. Comput. Mach., 25:81–91, 1978.
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 1992 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Codenotti, B., Leoncini, M., Resta, G. (1992). Repeated matrix squaring for the parallel solution of linear systems. In: Etiemble, D., Syre, JC. (eds) PARLE '92 Parallel Architectures and Languages Europe. PARLE 1992. Lecture Notes in Computer Science, vol 605. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-55599-4_120
Download citation
DOI: https://doi.org/10.1007/3-540-55599-4_120
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-55599-5
Online ISBN: 978-3-540-47250-6
eBook Packages: Springer Book Archive