Abstract
In a previous work we proposed two new Jacobi orderings for computing the eigenvalues of a symmetric matrix. The new orderings, combined with the application of the communication pipelining technique allow the exploitation of multi-port capabilities in hypercubes (sending several messages in parallel through the hypercube links). One of the proposed orderings, called Permuted-BR, is near optimal but only works for very large problems. The second ordering, called Degree-4, can be applied to any problem size but only reduces the communication cost by a factor of 4. In this paper we propose a new ordering, called Permuted-D4, that behaves like Permuted-BR ordering for big problems and like Degree-4 for small ones.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
W.C. Athas and C.L. Seitz. Multicomputers: message-passing concurrent computers. IEEE Computer, 99(7):9–24, August 1988.
C. Bischof. The two-sided jacobi method on a hypercube. In SIAM Proceedings of the Second Conference on Hypercube Multiprocessors, August 1987.
R.P. Brent and F.T. Luk. The solution of singular-value and symmetric eigenvalue problems on multiprocessor arrays. SIAM J. Sci. Statist. Comput., (6):69–84, August 1985.
A. González D. Royo and M. Valero-García. Jacobi orderings for multi-port hypercubes. In 12th International Paral.lel Processing Symposium and 9th Symposium on Patrallel and Distributed Processing, pages 88–98, 1998.
P.J. Eberlein. On one-sided jacobi methods for parallel computation. SIAM J. Algebraic Discrete Methods, (8):790–796,1987.
P.J. Eberlein and H. Park. Efficient implementation of jacobi algorithms and jacobi sets on distributed memory architectures. Journal of Parallel Distributed Computing, (8):358–366, August 1990.
G.R. Gao and S.J. Thomas. An optimal parallel jacobi-like solution method for singular value decomposition. In Proceedings of Int’l Conf. on Parallel Processing, pages 47–53, 1988.
A. Gonzàlez L. Díaz de Cerio and M. Valero-García. Communication pipelining in hypercubes. Parallel Processing Letters, 6(4), December 1996.
M. Lam. Software pipelining: An effective scheduling technique for vliw machines. In Conf. on Programming Language,Design and Implementation,pages 318–328,1988.
M. Mantharam and P.J. Eberlein. Block-recursive algorithm to generate jacobisets. Parallel Computing, (19):481–496, 1993.
M. Mantharam and P.J. Eberlein. New jacobi-sets for parallel computation. Parallel Computing, (19):437–454, 1993.
L.M. Ni and P.K. McKinley. A survey of wormhole routing techniques in directed networks. IEEE Computer, 26(2):62–76, February 1993.
J. Gotze S. Paul and M. Sauer. An efficient jacobi-like algorithm for parallel eigenvalue computation. IEEE Transactions on Computers, C42(9):1058–1063, 1993.
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 1999 Springer Science+Business Media New York
About this chapter
Cite this chapter
Royo, D., Valero-García, M., González, A. (1999). A New Jacobi Ordering for Multiple-Port Hypercubes. In: Yang, T. (eds) Parallel Numerical Computation with Applications. The Springer International Series in Engineering and Computer Science, vol 515. Springer, Boston, MA. https://doi.org/10.1007/978-1-4615-5205-5_6
Download citation
DOI: https://doi.org/10.1007/978-1-4615-5205-5_6
Publisher Name: Springer, Boston, MA
Print ISBN: 978-1-4613-7371-1
Online ISBN: 978-1-4615-5205-5
eBook Packages: Springer Book Archive