Abstract
The algebraic multigrain (AMG) is one of the most frequently used algorithms for the solution of large-scale sparse linear systems in many realistic simulations of science and engineering applications. However, as the concurrency of supercomputers increasing, the AMG solver increasingly leads to poor parallel scalability due to its coarse-level construction in the setup phase. In this paper, to improve the parallel scalability of the traditional AMG to solve the sequence of sparse linear systems arising from PDE-based simulations, we propose a new AMG procedure called αSetup-AMG based on an adaptive setup strategy. The main idea behind αSetup-AMG is the introduction of a setup condition in the coarsening process so that the setup is constructed as it needed instead of constructing in advance via an independent phase in the traditional procedure. As a result, αSetup-AMG requires fewer setup cost and level numbers for the sequence of linear systems. The numerical results on thousands of cores for a radiation hydrodynamics simulation in the inertial confinement fusion (ICF) application show the significant improvement in the efficiency of the αSetup-AMG solver.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
References
An, H., Mo, Z., Xu, X., Liu, X.: On choosing a nonlinear initial iterate for solving the 2-D 3-T heat conduction equations. J. Comput. Phys. 228(9), 3268–3287 (2009)
Baker, A., Falgout, R., Gamblin, T., Kolev, T., Schulz, M., Yang, U.: Scaling algebraic multigrid solvers: on the road to exascale. In: Bischof, C., Hegering, H.G., Nagel, W., Wittum, G. (eds.) Competence in High Performance Computing 2010, pp. 215–226. Springer, Berlin, Heidelberg (2010). https://doi.org/10.1007/978-3-642-24025-6_18
Baker, A., Gamblin, T., Schulz, M., Yang, U.: Challenges of scaling Algebraic Multigrid across modern multicore architectures. In: Proceedings of the 2011 IEEE International Parallel & Distributed Processing Symposium (IPDPS'11), pp. 275–286 (2011). https://doi.org/10.1109/IPDPS.2011.35
Baker, A., Klawonn, A., Kolev, T., Lanser, M., Rheinbach, O., Yang, U.: Scalability of classical algebraic multigrid for elasticity to half a million parallel tasks. In: Bungartz, H.J., Neumann, P., Nagel, W. (eds.) Software for Exascale Computing-SPPEXA 2013–2015. Lecture Notes in Computational Science and Engineering, vol. 113, pp. 113–140. Springer, Cham (2015). https://doi.org/10.1007/978-3-319-40528-5_6
Baldwin, C., Brown, P., Falgout, R., Graziani, F., Jones, J.: Iterative linear solvers in 2D radiation hydrodynamics code: methods and performance. J. Comput. Phys. 154, 1–40 (1999)
Ballard, G., Siefert, C., Hu, J.: Reducing communication costs for sparse matrix multiplication with algebraic multigrid. SIAM J. Sci. Comput. 38(3), C203–C231 (2016)
Bell, N., Dalton, S., Olson, L.: Exposing fine-grained parallelism in algebraic multigrid methods. SIAM J. Sci. Comput. 34, C123–C152 (2012)
Benzi, M.: Preconditioning techniques for large linear systems: a survey. J. Comput. Phy. 182, 418–477 (2002)
Bienz, A., Olson, L.: Reducing Network contention associated with parallel algebraic multigrid, SC14, 2 pages poster, New Orleans, USA (2014). http://sc14.supercomputing.org/sites/all/themes/sc14/files/archive/src_poster/poster_files/spost141s2-file2.pdf/. Accessed 5 May 2020
Bienz, A., Olson, L., Gropp, W.: Reducing parallel communication in algebraic multigrid with multi-step node aware communication, arXiv:1904.05838 (2019) https://arxiv.org/abs/1904.05838 Accessed 10 March 2020
Bienz, A., Falgout, R., Gropp, W., Olson, L., Schroder, J.: Reducing parallel communication in algebraic multigrid through sparsification. SIAM J. Sci. Comput. 38(5), S332–S357 (2016)
Brandt, A., McCormick, S., Ruge, J.: Algebraic multigrid for sparse matrix equations. In: Evans, D. (ed.) Sparsity and its application, pp. 257–284. Cambridge University Press, Cambridge (1984)
Falgout, R., Schroder, J.: Non-Galerkin coarse grids for algebraic multigrid. SIAM J. Sci. Comp. (36)3: C309-C334 (2014)
Fan, Z., Zhu, S., Pei, W., Ye, W., Li, M., Xu, X., Wu, J., Dai, Z., Wang, L.: Numerical inverstigation on the stabilization of the deceleration phase Rayleigh-Taylor instability due to alpha particle heating in ignition target. Euro. Phys. Lett. 99, 5003 (2012)
Gahvari, H.: Improving the performance and scalability of algebraic multigrid solvers through applied performance modeling. Ph.D. thesis, University of Illinois at Urbana-Champaign (2014)
Gahvari, H., Baker, A., Schulz, M., Yang, U., Jordan, K., Gropp, W.: Modeling the performance of an algebraic multigrid cycle on HPC platforms. In Lowenthal, D., Supinski, B., McKee, S.(eds.) Proceedings of the 25th International Conference on Supercomputing, pp.172–181, ACM, Tucson, USA (2011)
Gu, T., Xu, X., Liu, X., An, H., Hang, X.: Iterative methods and preconditioning techniques. Information and computation science series no. 76. Science Press, Beijing (2015)
Henson, V., Yang, U.: BoomerAMG: a parallel algebraic multigrid solver and preconditioner. Appl. Numer. Math. 41, 155–177 (2002)
Hypre. https://computation.llnl.gov/projects/hypre. Accessed 10 March 2020. (2019)
JASMIN. https://www.caep-scns.ac.cn/JASMIN.php (2019) Accessed 10 March 2020.
Jin, C., Cai, X.-C.: A preconditioned recycling GMRES solver for stochastic Helmholtz problems. CiCP 6(2), 342–353 (2009)
JXPAMG. https://www.multigrid.org/solver/jxpamg.html (2019) Accessed 10 March 2020.
Lu, Y.: Paving the way for China exascale computing. CCF Trans High Perform Comput 1, 63–72 (2019)
Meijerink, J., van der Vorst, H.: An iterative solution method for linear systems of which the coefficient matrix is a symmetric M-matrix. Math. Comput. 137, 148–162 (1977)
Mo, Z., Xu, X.: Relaxed RS0 or CLJP coarsening strategy for parallel AMG methods. Parallel Comput. 33(3), 174–185 (2007)
Mo, Z., Zhang, A., Cao, X., Liu, Q., Xu, X., An, H., Pei, W., Zhu, S.: JASMIN: a parallel software infrastructure for scientific computing. Front. Comput. Sci. China 4(4), 480–488 (2010)
Parks, M., de Sturler, E., Mackey, G., Johnson, D., Maiti, S.: Recycling Krylov subspaces for sequences of linear systems. SIAM J. Sci. Comput. 28, 1651–1674 (2006)
Pei, W.: The construction of simulation algorithm for laser fusion. CiCP 2(2), 255–270 (2007)
Pei, W., Zhu, S.: Scientific computing in laser fusion. Physics 38(8), 559–568 (2009)
Ruge, J., Stüben, K.: Algebraic multigrid (AMG). In: McCormick, S. (ed.) Multigrid Methods, Frontiers in Applied Mathematics, vol. 3, pp. 73–130. SIAM, Philadelphia (1987)
Saad, Y.: Iterative methods for sparse linear systems. SIAM, Philadelphia (2003)
De Sterck, H., Yang, U., Heys, J.: Reducing complexity in parallel algebraic multigrid preconditioners. SIAM J. Mat. Anal. Appl. 27(4), 1019–1039 (2006)
Stüben, K.: Algebraic multigrid (AMG): an introduction with applications. In: Trottenberg, U., Oosterlee, C., Schuller, A. (eds.) Multigrid, pp. 413–532. Academic Press, New York (2001)
Toselli, A., Widlund, O.: Domain decomposition methods: algorithms and theory. Spring, Berlin (2005)
Trottenberg, U., Oosterlee, C., Schuller, A.: Multigrid. Academic Press, Singapore (2001)
Vassilevski, P., Yang, U.: Reducing communication in algebraic multigrid using additive variants. Numer. Linear Algebra Appl. 21, 275–296 (2014)
Xu, X.: Parallel algebraic multigrid methods: state-of-the art and challenges for extreme-scale applications. J. Numer. Methods Comput. Appl. 40(4), 243–260 (2019)
Xu, X., Mo, Z.: Scalability analysis for parallel algebraic multigrid algorithm. Chin. J. Comput. Phys. 24(4), 387–394 (2007)
Xu, X., Mo, Z.: Algebraic interface based coarsening AMG preconditioner for multi-scale sparse matrices with applications to radiation hydrodynamics computation. Numer Linear Algebra Appl. 24(2), e2078 (2017). https://doi.org/10.1002/nla.2078
Xu, X., Mo, Z., An, H.: An algebraic two-level method for 2D–3T radiation diffusion equations. Chin. J. Comput. Phys. 26(1), 1–8 (2009)
Xu, X., Mo, Z., Wu, L.: Analysis of communication-to computation based on aysmptotic size for iterative methods. Chin. J. Comput. 36(4), 782–789 (2013)
Xu, X., Mo, Z., An, H.: An adaptive AMG preconditioning strategy for solving large-scale sparse linear systems. Sci. Sinica Inform. 46(10), 1411–1420 (2016)
Yang, U.: Parallel algebraic multigrid methods - high performance preconditioners. In: Bruaset, A., Tveito, A. (eds.) Numerical Solution of PDEs on Parallel Computers. Lecture Notes in Computational Science and Engineering, vol. 51, pp. 209–236. Springer, Cham (2006). https://doi.org/10.1007/3-540-31619-1_6
Yue, X., Shu, S., Xu, X., Zhou, Z.: An adaptive combined preconditioner with applications in radiation diffusion equations. CiCP 18(5), 1313–1335 (2015)
Zhou, Z., Xu, X., Shu, S., Feng, C., Mo, Z.: An adaptive two-level preconditioner for 2D–3T radiation diffusion equations. Chin. J. Comput. Phys. 29(4), 475–548 (2012)
Acknowledgements
The authors would like to thank the reviewers of this paper for their constructive and helpful comments and suggestions. This work is supported by National Key R&D Program of China under Grant No. 2017YFB0202103, Science Challenge Project under Grant no. TZZT2016002, and National Nature Science Foundation of China under Grant Nos. 61370067 and 11971414.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
On behalf of all authors, the corresponding author states that there is no conflict of interest.
Rights and permissions
About this article
Cite this article
Xu, X., Mo, Z., Yue, X. et al. αSetup-AMG: an adaptive-setup-based parallel AMG solver for sequence of sparse linear systems. CCF Trans. HPC 2, 98–110 (2020). https://doi.org/10.1007/s42514-020-00033-w
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s42514-020-00033-w