OpenMP Thread Affinity for Matrix Factorization on Multicore Systems
Beata Bylina, Jarosław Bylina
DOI: http://dx.doi.org/10.15439/2017F231
Citation: Proceedings of the 2017 Federated Conference on Computer Science and Information Systems, M. Ganzha, L. Maciaszek, M. Paprzycki (eds). ACSIS, Vol. 11, pages 489–492 (2017)
Abstract. The aim of this paper is to investigate the impact of thread affinity on computing performance for matrix factorization on shared memory multicore systems with hierarchical memory. We consider two parallel block matrix factorizations (LU and WZ) and employ thread affinity to improve their performance. We study decomposition without pivoting and we compare differences between various affinity strategies for diagonally dominant matrices. Our results show that the choice of thread affinity has the measurable impact on the performance of the matrice factorizations.
References
- Matthias Diener, Eduardo H. M. Cruz, Marco A. Z. Alves, Philippe O. A. Navaux, and Israel Koren. Affinity-based thread and data mapping in shared memory systems. ACM Comput. Surv., 49(4):64:1–64:38, December 2016.
- J. Dongarra, J. DuCroz, I. S. Duff, and S. Hammarling. A set of level-3 Basic Linear Algebra Subprograms. ACM Trans. Math. Software, 16:1–28, 1990.
- D.J. Evans and M. Hatzopoulos. A parallel linear system solver. International Journal of Computer Mathematics, 7(3):227–238, 1979.
- P. Yalamov and D.J. Evans. The WZ matrix factorisation method. Parallel Computing, 21(7):1111–1120, 1995.