Abstract
In this paper we study the parallel scalability of variants of an algebraic additive Schwarz preconditioner for the solution of large three dimensional convection diffusion problems in a non-overlapping domain decomposition framework. To alleviate the computational cost, both in terms of memory and floating-point complexity, we investigate variants based on a sparse approximation or on mixed 32- and 64-bit calculation. The robustness and the scalability of the preconditioners are investigated through extensive parallel experiments on up to 2,000 processors. Their efficiency from a numerical and parallel performance view point are reported.
Similar content being viewed by others
References
Amestoy, P.R., Duff, I.S., Koster, J., L’Excellent, J.-Y.: A fully asynchronous multifrontal solver using distributed dynamic scheduling. SIAM J. Matrix Anal. Appl. 23(1), 15–41 (2001)
Amestoy, P.R., Guermouche, A., L’Excellent, J.-Y., Pralet, S.: Hybrid scheduling for the parallel solution of linear systems. Parallel Comput. 32(2), 136–156 (2006)
Anderson, E., Bai, Z., Bischof, C., Demmel, J., Dongarra, J., Du Croz, J., Greenbaum, A., Hammarling, S., McKenney, A., Ostrouchov, S., Sorenson, D.: LAPACK Users’ Guide, 2nd edn. Society for Industrial and Applied Mathematics, Philadelphia, PA, USA (1995)
Arioli, M., Duff, I.S., Gratton, S., Pralet, S.: Note on GMRES preconditioned by a perturbed LDLt decomposition with static pivoting. SIAM J. Sci. Comput. 25(5), 2024–2044 (2007)
Buttari, A., Dongarra, J., Kurzak, J., Luszczek, P., and Tomov, S.: Using mixed precision for sparse matrix computations to enhance the performance while achieving 64-bit accuracy. ACM Trans. Math. Software 34(4), 1–22 (2008)
Cai, X.C., Saad, Y.: Overlapping domain decomposition algorithms for general sparse matrices. Numer. Linear Algebra Appl. 3, 221–237 (1996)
Carvalho, L.M., Giraud, L., Meurant, G.: Local preconditioners for two-level non-overlapping domain decomposition methods. Numer. Linear Algebra Appl. 8(4), 207–227 (2001)
Drkošová, J., Rozložník, M., Strakoš, Z., Greenbaum, A.: Numerical stability of the GMRES method. BIT, 35, 309–330 (1995)
Duff, I.S., Reid, J.K.: The multifrontal solution of indefinite sparse symmetric linear systems. ACM Trans. Math. Software 9, 302–325 (1983)
Frayssé, V., Giraud, L., Gratton, S.: Algorithm 881: a set of FGMRES routines for real and complex arithmetics on high performance computers. ACM Trans. Math. Software 35(2), 1–12 (2008)
Frayssé, V., Giraud, L., Gratton, S., Langou, J.: Algorithm 842: a set of GMRES routines for real and complex arithmetics on high performance computers. ACM Trans. Math. Software 31(2), 228–238 (2005)
Gale, A.C., Almeida, R.C., Malta, S.M.C., Loula, A.F.D.: Finite element analysis of convection dominated reaction-diffusion problems. Appl. Numer. Math. 48(2), 205–222 (2004)
Giraud, L., Gratton, S., Langou, J.: Convergence in backward error of relaxed GMRES. SIAM J. Sci. Comput. 29(2), 710–728 (2007)
Giraud, L., Haidar, A., Watson, L.T.: Parallel scalability study of hybrid preconditioners in three dimensions. Parallel Comput. 34, 363–379 (2008)
Giraud, L., Marrocco, A., Rioual, J.-C.: Iterative versus direct parallel substructuring methods in semiconductor device modelling. Numer. Linear Algebra Appl. 12(1), 33–53 (2005)
Haidar, A.: On the parallel scalability of hybrid solvers for large 3D problems. Ph.D. dissertation, INPT, June 2008. TH/PA/08/57.
Kurzak, J., Dongarra, J.: Implementation of the mixed-precision high performance LINPACK benchmark on the CELL processor. Technical Report LAPACK Working Note #177 UT-CS-06-580, University of Tennessee Computer Science (2006) September
Langou, J., Langou, J., Luszczek, P., Kurzak, J., Buttari, A., Dongarra, J.: Exploiting the performance of 32 bit floating point arithmetic in obtaining 64 bit accuracy. Technical Report LAPACK Working Note #175 UT-CS-06-574, University of Tennessee Computer Science (2006) April
Mathew, T.: Domain Decomposition Methods for the Numerical Solution of Partial Differential Equations. Springer, New York (2008)
Morton, K.W.: Numerical Solution of Convection-Diffusion Problems. Chapman & Hall, London (1996)
Paige, C., Rozložník, M., Strakoš, Z.: Modified Gram-Schmidt (MGS), least-squares, and backward stability of MGS-GMRES. SIAM J. Matrix Anal. Appl. 28(1), 264–284 (2006)
Quarteroni, A., Valli, A.: Domain Decomposition Methods for Partial Differential Equations. Numerical mathematics and scientific computation. Oxford Science Publications, Oxford (1999)
Radicati, G., Robert, Y.: Parallel conjugate gradient-like algorithms for solving nonsymmetric linear systems on a vector multiprocessor. Parallel Comput. 11, 223–239 (1989)
Roos, H.-G., Stynes, M., Tobiska, L.: Numerical methods for singularly perturbed differential equations. Springer Series in Computational Mathematics, vol. 24. Springer-Verlag, Berlin. Convection-diffusion and flow problems (1996)
Saad, Y.: A flexible inner-outer preconditioned GMRES algorithm. SIAM J. Sci. Comput. 14, 461–469 (1993)
Saad, Y., Schultz, M.H.: GMRES: a generalized minimal residual algorithm for solving nonsymmetric linear systems. SIAM J. Sci. Stat. Comp. 7, 856–869 (1986)
Shanthi, V., Ramanujam, N.: Asymptotic numerical methods for singularly perturbed fourth order ordinary differential equations of convection-diffusion type. Appl. Math. Comput. 133(2–3), 559–579 (2002)
Smith, B.F., Bjørstad, P.E., Gropp, W.: Domain Decomposition: Parallel Multilevel Methods for Elliptic Partial Differential Equations. Cambridge University Press, Cambridge (1996)
Stynes, M.: Steady-state convection-diffusion problems. Acta Numer. 14, 445–508 (2005)
Toselli, A., Widlund, O.: Domain Decomposition Methods—Algorithms and Theory. Springer, Berlin (2005)
Author information
Authors and Affiliations
Corresponding author
Additional information
This research activity was partially supported within the framework of the ANR-CIS project Solstice (ANR-06-CIS6- 010).
Rights and permissions
About this article
Cite this article
Giraud, L., Haidar, A. Parallel algebraic hybrid solvers for large 3D convection-diffusion problems. Numer Algor 51, 151–177 (2009). https://doi.org/10.1007/s11075-008-9248-x
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11075-008-9248-x