Abstract
While ILU preconditioner is a powerful and popular preconditioning method for Krylov iterative solvers on sparse matrices derived from finite element analysis, it have been exploerd the scalable hybrid parallelization scheme for ILU preconditioner targetting multi/many-core clusters. Hierarchical Interface Decompostion (HID) is a robust and efficient parallel method for ILU preconditioner. The extended version of HID (ExHID), our proposed method, introduces thicker level-2 connector in order to consider fill-ins. Basing on HID and ExHID we developed hybrid parallel ILU preconditioner with fill-ins using OpenMP/MPI hybrid parallel programing models. While inter-node parallelization is based on HID/ExHID, we applied two different methods, multicolor based reordering and HID/ExHID to intra-node parallelization. The two implementations according to different hybrid strategy, HID(inter-node)-HID (intra-node) and HID(inter-node)-MC(intra-node), are evaluated through strong scaling tests and the better hybrid strategy is explored. HID-HID generally results with better convergence and less fill-ins. On the other hand, HID-MC could be more stable strategy than HID-HID when increasing the number of threads per process.
Access provided by Autonomous University of Puebla. Download to read the full chapter text
Chapter PDF
Similar content being viewed by others
Keywords
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.
References
Jones, M.T., Plassmann, P.E.: Scalable iterative solution of sparse linear systems. Parallel Computing 20(5), 753–773 (1994)
Saad, Y., Sosonkina, M.: Distributed schur complement techniques for general sparse linear systems. SIAM Journal on Scientific Computing 21(4), 1337–1356 (2000)
Parallel iterative solvers of geofem with selective blocking preconditioning for nonlinear contact problems on the earth simulator (2003)
The Impact of Parallel Programming Models on the Linear Algebra Performance for Finite Element Simulations (2007)
Washio, T., Hisada, T., Watanabe, H., Tezduyar, T.E.: A robust preconditioner for fluid-structure interaction problems. Computer Methods in Applied Mechanics and Engineering 194(39-41), 4027–4047 (2005)
Nakajima, K.: Parallel preconditioning methods with selective fill-ins and selective overlapping for ill-conditioned problems in finite-element methods. In: Shi, Y., van Albada, G.D., Dongarra, J., Sloot, P.M.A. (eds.) ICCS 2007, Part III. LNCS, vol. 4489, pp. 1085–1092. Springer, Heidelberg (2007)
Saad, Y., Sosonkina, M.: Enhanced parallel multicolor preconditioning techniques for linear systems. UMSI Research Report/University of Minnesota (Minneapolis, Mn). Supercomputer Institute, vol. 99, p. 4 (1999)
Henon, P., Saad, Y., et al.: A parallel multistage ilu factorization based on a hierarchical graph decomposition. SIAM Journal on Scientific Computing 28(6), 2266 (2006)
Nakajima, K.: Flat mpi vs. hybrid: Evaluation of parallel programming models for preconditioned iterative solvers on “t2k open supercomputer”. In: International Conference on Parallel Processing Workshops, ICPPW 2009, pp. 73–80 (2009)
Mattson, T., Sanders, B., Massingill, B.: Patterns for parallel programming. Addison-Wesley Professional (2004)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2013 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Hayashi, M., Nakajima, K. (2013). OpenMP/MPI Hybrid Parallel ILU(k) Preconditioner for FEM Based on Extended Hierarchical Interface Decomposition for Multi-core Clusters. In: Daydé, M., Marques, O., Nakajima, K. (eds) High Performance Computing for Computational Science - VECPAR 2012. VECPAR 2012. Lecture Notes in Computer Science, vol 7851. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-38718-0_28
Download citation
DOI: https://doi.org/10.1007/978-3-642-38718-0_28
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-38717-3
Online ISBN: 978-3-642-38718-0
eBook Packages: Computer ScienceComputer Science (R0)