Abstract
We present the results of a study of the parallel algorithms based on MPI and OpenMP for vector splitting schemes in heat transfer problems. We compare the three parallel implementations: MPI, “simple” MPI/OpenMP (#pragma omp directives applied to MPI-based code), and MPI/OpenMP with “postman” threads. The main idea of the last algorithm is to assign one thread within each computational node to perform the data transfer. This approach allows us to implement overlapping of useful computations and data transfer. The results show that the introducing postman threads can significantly improve the performance of an MPI/OpenMP implementation; nevertheless, for the considered class of numerical algorithms, it is more reasonable to use an MPI implementation.
Similar content being viewed by others
References
I. Mitin, A. Kalinkin, and Yu. Laevsky, “A Parallel Iterative Solver for Positive-Definite Systems with Hybrid MPI-OpenMP Parallelization for Multi-Core Clusters,” J. Comput. Sci. 6(3), 463–468 (2012).
T. A. Kandryukova, “On Numerical Studying a Problem of Filtration Gas Combustion with Computer Systems with Shared Memory,” in Proceedings of the Young Scientists Conference of the Institute of Computational Mathematics and Mathematical Geophysics, 2012, URL: http://parbz.sscc.ru/fcp/kmu2012/Kandryukova.pdf
R. Rabenseifner, G. Hager, and G. Jost, “Hybrid MPI/OpenMP Parallel Programming on Clusters of Multi-Core SMP Nodes,” in Proceedings of the 17 Euromicro International Conference on Parallel, Distributed, and Network-based Processing (Weimar, 2009), pp. 427–436.
A. Kalinkin, “Intel Direct Sparse Solver for Clusters, a Research Project for Solving Large Sparse Systems of Linear Algebraic Equations on Clusters,” URL: http://www.cerfacs.fr/files/cerfacs-algo/conferences/PastWorkshops/SparseDays2013/Kalinkin.pdf
H. He, “Hybrid OpenMP and MPI Programming and Tuning,” URL: http://www.nersc.gov/assets/NUGMeetings/2004/NUG2004yhehybrid.ppt
K. V. Voronin and Yu. M. Laevsky, “On Splitting Schemes in the Mixed Finite Element Method,” Numer. Anal. Appl. 5(2), 150–155 (2012).
K. V. Voronin and Yu. M. Laevsky, “Splitting Schemes in the Mixed Finite-Element Method for the Solution of Heat Transfer Problems,” Math. Models Comput. Simulations 5(2), 167–174 (2013).
V. A. Vernikovsky, A. E. Vernikovsky, O. P. Polyansky, Yu. M. Laevsky, N. Yu. Matushkin, K. V. Voronin, “A Tectonormal Model for the Formation of an Orogen at the Post-Collisional Stage (by the Example of the Yenisei Ridge, Eastern Siberia),” Russian geology and geophysics No. 52, 24–39 (2011).
A.E. Vernikovskaya, V.M. Datsenko, V. A. Vernikovsky, N. Yu. Matushkin, Yu. M. Laevsky, I. V. Romanova, A.V. Travin, K. V. Voronin, and E. N. Lepekhina, “Magmatism Evolution and Carbonatitegranite Association in the Neoproterozoic Active Continental Margin of the Siberian Craton: Thermochronological Reconstructions,” Dokl. Earth Sci. 448(2), 161–167 (2013).
P. A. Raviart and J. M. Thomas, “A Mixed Finite Element Method for 2nd Order Elliptic Problems,” in Mathematical Aspects of Finite Element Methods: Lecture Notes in Mathematics, Vol. 606 (Springer, Berlin, 1977), pp. 292–315.
J. Douglas and J. E. Gunn, “A General Formulation of Alternating Direction Methods,” Numer. Math. 6, 428–453 (1964).
Siberian Supercomputer Center (SSCC), URL: http://www2.sscc.ru
Author information
Authors and Affiliations
Corresponding author
Additional information
Original Russian Text © K.V. Voronin, 2014, published in Sibirskii Zhurnal Industrial’noi Matematiki, 2014, Vol. XVII, No. 2, pp. 41–49.
Rights and permissions
About this article
Cite this article
Voronin, K.V. A numerical study of an MPI/OpenMP implementation based on asynchronous threads for a three-dimensional splitting scheme in heat transfer problems. J. Appl. Ind. Math. 8, 436–443 (2014). https://doi.org/10.1134/S199047891403017X
Received:
Published:
Issue Date:
DOI: https://doi.org/10.1134/S199047891403017X