Abstract
In this paper, one new parallel version of GPBi-CG method (PGPBi-CG method, in brief) is proposed for solving large sparse linear systems with unsymmetrical coefficient matrices on distributed parallel environments. The method reduces three global synchronization points to one by reconstructing GPBi-CG method and the communication time required for the inner product can be efficiently overlapped with computation time of vector updates. It combines the elements of numerical stability with the characteristics of design of parallel algorithms. The cost is only slightly increased computation time, which can be ignored, compared with the reduction of communication time. Performance and isoefficiency analysis shows that the PGPBi-CG method has better parallelism and scalability than the GPBi-CG method. Numerical experiments show that the scalability can be improved by a factor 3 and the improvement in parallel communication performance approaches 66.7 %.
Similar content being viewed by others
References
Bhaya A, Bliman PA, Niedu G, Pazos F (2012) A cooperative conjugate gradient method for linear systems permitting multithread implementation of low complexity. In: Proceedings of the IEEE Conference on Decision and Control, pp. 638ff
Bücker HM, Sauren M (1996) A parallel version of the quasi-minimal residual method based on coupled two-term recurrences. In: Proceedings of Workshop on Applied Parallel Computing in Industrial Problems and Optimization (Para96). Technical University of Denmark, Lyngby, Denmark, Springer-Verlag, August 1996
Collignon TP, vanGijzen MB (2010) Fast solution of nonsymmetric linear systems on Grid computers using parallel variants of IDR(\(s\)). Delft University of Technology, Reports of the Department of Applied Mathematical Analysis, Report 10–05
de Sturler E, van der Vorst HA (1995) Reducing the effect of the global communication in GMRES(m) and CG on parallel distributed memory computers. Appl Numer Math 18:441–459
de Sturler E (1996) A performance model for Krylov subspace methods on mesh-based parallel computers. Parallel Comput 22:57–74
Freund RW, Nachtigal NM (1991) QMR: a quasi-minimal residual method for non-Hermitian linear systems. Numer Math 60:315–339
Grigori L, Moufawad S (2013) Communication avoiding ILU0 preconditioner. http://hal.inria.fr/docs/00/80/32/50/PDF/RR-8266
Gu T-X, Liu X-P, Mo Z-Y (2004) Multiple search direction conjugate gradient method I: methods and their propositions. Int J Comput Math 81(9):1133–1143
Gu T-X, Liu X-P, Mo Z-Y (2004) Multiple search direction conjugate gradient method II: theory and numerical experiments. Int J Comput Math 81(10):1289–1307
Gu T-X, Zuo X-Y, Zhang L-T, Zhang W-Q, Sheng Z-Q (2007) An improved bi-conjugate residual algorithm suitable for distributed parallel computing. Appl Math Comput 186:1243–1253
Gu T-X Zuo X-Y, Liu X-P, Li P-L (2009) An improved parallel hybrid bi-conjugate gradient method suitable for distributed parallel computing. J Comput Appl Math 226:56–65
Liu J, Liu X-P, Chi L-H, Hu Q-F, Li X-M (2005) An improved conjugate residual algorithm for large symmetric linear systems. In: Proceedings of the Joint Conference of ICCP6 and CCP2003, Comput. Phys., Xi-Jun Yu (ed.) Rinton Press, New Jersey, USA, pp. 325C328
Liu X-P, Gu T-X, Hang X-D, Sheng Z-Q (2006) A parallel version of QMRCGSTAB method for large linear systems in distributed parallel environments. Appl Math Comput 172(2):744C752
Saad Y (1996) Iterative methods for sparse linear systems. PWS Publishing Company, Boston
Sonneveld P, van Gijzen M (2008) IDR(s): a family of simple and fast algorithms for solving large nonsymmetric systems of linear equations. SIAM J Sci Comput 31(2):1035C1062
Sogabe T, Zhang S-L (2003) Extended conjugate residual methods for solving nonsymmetric linear systems. In: Yuan Y-X (ed) Numerical linear algebra and optimization. Science Press, Beijing/New York, pp. 88–99
van der Vorst HA (1992) Bi-CGSTAB: a fast and smoothly converging variant of bi-CG for the solution of nonsymmetric linear systems. SIAM J Sci Stat Comput 13:631–644
Yang TR, Lin HX (1997) The improved quasi-minimal residual method on massively distributed memory computers. In: Proceedings of The International Conference on High Performance Computing and Networking (HPCN-97)
Yang TR (2002) The improved CGS method for large and sparse linear systems on bulk synchronous parallel architectures. In: In 5th International Conference on Algorithms and Architectures for Parallel Processing, IEEE Computer Society, pp. 232–237
Yang TR, Brent RP (2002) The improved BICGSTAB method for large and sparse unsymmetric linear systems on parallel distributed memory architectures. In: 5th International Conference on Algorithms and Architectures for Parallel Processing, IEEE Computer Society, pp. 324–328
Yang TR, Brent RP (2003) The improved BiCG method for large and sparse linear systems on parallel distributed memory architectures. Inf J 6:349–360
Zhang S-L (1997) GPBi-CG: generalized product-type methods based on Bi-CG for solving nonsymmetric linear systems. SIAM J Sci Comput 18:537–551
Acknowledgments
The authors would like to thank the referees and Editor for their helpful and detailed suggestions for revising this manuscript.
Author information
Authors and Affiliations
Corresponding authors
Additional information
Communicated by Philippe Devloo.
This research of this author is supported by the NSFC (61202098, 61203179, 61170309, 11171039, 11471098 and 61272544), China Postdoctoral Science Foundation (2014M552001), Henan Province Postdoctoral Science Foundation (2013031), Aeronautical Science Foundation of China (2013ZD55006), Project of Youth Backbone Teachers of Colleges and Universities in Henan Province (2013GGJS-142), Major Project of development foundation of science and technology of CAEP (2012A0202008).
Rights and permissions
About this article
Cite this article
Zuo, XY., Zhang, LT., Gu, TX. et al. A parallel version of GPBi-CG method suitable for distributed parallel computing. Comp. Appl. Math. 35, 579–593 (2016). https://doi.org/10.1007/s40314-014-0206-z
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s40314-014-0206-z
Keywords
- PGPBi-CG method
- Krylov subspace
- Sparse unsymmetrical linear systems
- Global communication
- Distributed parallel environments