Abstract
This paper presents numerical experiments with assorted versions of parallel LU matrix decomposition algorithms (Gauss and Crout algorithm). The tests have been carried out on the hardware platform with fourcore Skylake processor featuring hyperthreading technology doubling virtually core number. Parallelization algorithms have been implemented with the aid of classic POSIX threads library. Experiments have shown that basic 4-thread acceleration of all parallel implementations is almost equal to the number of threads/processors. Both algorithms are worth considering in real-world applications (Florida University collection). Gauss algorithm is a better performer, with respect to timing, in the case of matrices with lower density of nonzeros, as opposed to higher density matrices. The latter are processed more efficiently with the aid of Crout algorithm implementation.
References
[1] HighamN. J.,Gaussian elimination (Advanced Review), In: WIREs Computational Statistics, 2011, 3(3), 230-23810.1002/wics.164Search in Google Scholar
[2] Mayer J., A multilevel Crout ILU preconditioner with pivoting and row permutation, Numerical Linear Algebra with Applications, 2007, 14(10), 771-78910.1002/nla.554Search in Google Scholar
[3] Sibai F. N., Performance modeling and analysis of parallel Gaussian elimination on multi-core computers, Journal of King Saud University - Computer and Information Sciences, 2014, 26, 41- 5410.1016/j.jksuci.2013.03.002Search in Google Scholar
[4] Zhang L., Sun Y., Ma J., A parallel Crout algorithm based on TBB, In: Proceedings of the 2nd International Conference on Software Engineering and Service Science, IEEE, 2011, 239-24210.1109/ICSESS.2011.5982298Search in Google Scholar
[5] Demmel J. W., Eisenstat S. C., Gilbert J., Li X., Liu J. W. H., Technical Report UCB//CSD-95-883, Computer Science Division, U.C. Berkeley, 1995Search in Google Scholar
[6] Donfack S., Dongarra J., Faverge M., Gates M., Kurzak J., Łuszczek P., Yamazaki I., On algorithmic variants of parallel Gaussian elimination: comparison of implementations in terms of performance and numerical properties, Research Report, 2013, https://hal.inria.fr/hal-00867837Search in Google Scholar
[7] Huckle T., Parallel numerics, Technische Universität München, Institut für Informatik, 2006Search in Google Scholar
[8] Butenhof D. R., Programming with POSIX Threads, Addison Wesley, Boston, 1997Search in Google Scholar
[9] Kaya D.,Wright K., Parallel algorithms for LU decomposition on a shared memory multiprocessor, AppliedMathematics and Computation, 2005, 163(1), 179-191.10.1016/j.amc.2004.01.027Search in Google Scholar
© by Marek Stabrowski, published by De Gruyter
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License.
