PPAM 2007: Parallel Processing and Applied Mathematics pp 1257-1264 | Cite as
Scalability and Performance Analysis of a Probabilistic Domain Decomposition Method
Abstract
In this paper, we analyze the scalability and performance of a probabilistic domain decomposition strategy for solving linear elliptic boundary-value problems. Such a strategy consists of a hybrid numerical scheme based on a probabilistic method along with a domain decomposition, and full decoupling can be accomplished. It is shown that such a method performs well for an arbitrarily large number of processors, while the classical deterministic approach is strongly affected by intercommunications. Therefore, the overall performance degrades dramatically for rather large number of processors. Furthermore, we find that the probabilistic method is scalable as the number of subdomains, i.e., the number of processors involved, increases. This fact is clearly illustrated by an example.
Preview
Unable to display preview. Download preview PDF.
References
- 1.Acebrón, J.A., Busico, M.P., Lanucara, P., Spigler, R.: Domain decomposition solution of elliptic boundary-value problems via Monte Carlo and quasi-Monte Carlo methods. SIAM J. Sci. Comput. 27, 440–457 (2005)CrossRefMathSciNetMATHGoogle Scholar
- 2.Acebrón, J.A., Busico, M.P., Lanucara, P., Spigler, R.: Probabilistically induced domain decomposition methods for elliptic boundary-value problems. J. Comput. Phys. 210, 421–438 (2005)CrossRefMathSciNetMATHGoogle Scholar
- 3.Acebrón, J.A., Spigler, R.: Fast simulations of stochastic dynamical systems. J. Comput. Phys. 208, 106–115 (2005)CrossRefMathSciNetMATHGoogle Scholar
- 4.Acebrón, J.A., Spigler, R.: Supercomputing applications to the numerical modeling of industrial and applied mathematics problems. J. Supercomputing 40, 67–80 (2007)CrossRefGoogle Scholar
- 5.Acebrón, J.A., Durán, R., Rico, R., Spigler, R.: A new domain decomposition approach suited for grid computing. LNCS. Springer, Heidelberg (in press, 2007)Google Scholar
- 6.Caflisch, R.E.: Monte Carlo and quasi Monte Carlo methods. In: Acta Numerica, pp. 1–49. Cambridge University Press, Cambridge (1998)Google Scholar
- 7.Freidlin, M.: Functional integration and partial differential equations. Annals of Mathematics Studies no. 109. Princeton Univ. Press, Princeton (1985)MATHGoogle Scholar
- 8.Geist, G.A.: Progress towards Petascale Virtual Machines. In: Dongarra, J., Laforenza, D., Orlando, S. (eds.) EuroPVM/MPI 2003. LNCS, vol. 2840, pp. 10–14. Springer, Heidelberg (2003)Google Scholar
- 9.Jansons, K.M., Lythe, G.D.: Exponential timestepping with boundary test for stochastic differential equations. SIAM J. Sci. Comput. 24, 1809–1822 (2003)CrossRefMathSciNetMATHGoogle Scholar
- 10.Keyes, D.E.: How scalable is domain decomposition in practice? In: The Eleventh International Conference on Domain Decomposition Methods (London, 1998) (electronic), DDM.org, Augsburg, pp. 286–297 (1999)Google Scholar
- 11.Karypis, G., Kumar, V.: ParMETIS–parallel graph partitioning and field–reducing matrix ordering, http://glaros.dtc.umn.edu/gkhome/metis/parmetis/overview
- 12.Niederreiter, H.: Random number generation and quasi Monte-Carlo methods. SIAM, Philadelphia (1992)MATHGoogle Scholar
- 13.Quarteroni, A., Valli, A.: Domain decomposition methods for partial differential equations. Oxford Science Publications, Clarendon Press (1999)Google Scholar
- 14.Toselli, A., Widlund, O.: Domain Decomposition Methods - Algorithms and Theory. Springer Series in Computational Mathematics, vol. 34. Springer, Heidelberg (2005)MATHGoogle Scholar
- 15.Li, Z., Saad, Y., Sosonkina, M.: pARMS: a parallel version of the algebraic recursive multilevel solver. Numerical Linear Algebra with Applications 10, 485–509 (2003)CrossRefMathSciNetMATHGoogle Scholar