Abstract
In this article, we study the effects of network topology and load balancing on the performance of a new parallel algorithm for solving triangular systems of linear equations on distributed-memory message-passing multiprocessors. The proposed algorithm employs novel runtime data mapping and workload redistribution methods on a communication network which is configured as a toroidal mesh. A fully parameterized theoretical model is used to predict communication behaviors of the proposed algorithm relevant to load balancing, and the analytical performance results correctly determine the optimal dimensions of the toroidal mesh, which vary with the problem size, the number of available processors, and the hardware parameters of the machine. Further enhancement to the proposed algorithm is then achieved through redistributing the arithmetic workload at runtime. Our FORTRAN implementation of the proposed algorithm as well as its enhanced version has been tested on an Intel iPSC/2 hypercube, and the same code is also suitable for executing the algorithm on the iPSC/860 hypercube and the Intel Paragon mesh multiprocessor. The actual timing results support our theoretical findings, and they both confirm the very significant impact a network topology chosen at runtime can have on the computational load distribution, the communication behaviors and the overall performance of parallel algorithms.
Similar content being viewed by others
References
R. H. Bisseling and J. G. G. van de Vorst. Parallel triangular system solving on a mesh network of transputers. SIAM J. Sci. Stat. Comput., 12:787—799, 1991.
T. H. Dunigan. Performance of the Intel iPSC/860 Hypercube. Technical Report ORNL/TM-11491, Oak Ridge National Laboratory, Oak Ridge, Tennessee, 1990.
T. H. Dunigan. Communication performances of the Intel Touchstone DELTA Mesh. Technical Report ORNL/TM-11983, Oak Ridge National Laboratory, Oak Ridge, Tennessee, 1992.
M. T. Heath and C. H. Romine. Parallel solution of triangular systems on distributed-memory mul-tiprocessors. SIAM J. Sci. Stat. Comput., 9:558—588, 1988.
G. Li and T. F. Coleman. A parallel triangular solver on a distributed-memory multiprocessor. SIAM J. Sci. Stat. Comput., 9:485—502, 1988.
G. Li and T. F. Coleman. A new method for solving triangular system on distributed-memory mul-tiprocessors. SIAM J. Sci. Stat. Comput., 10:382—396, 1989.
L. D. J. C. Loyens and R. H. Bisseling. The formal construction of a parallel triangular system solver. In Lecture Notes in Computer Science, Number 375, pages 325—334. Springer-Verlag, 1989.
C. H. Romine and J. M. Ortega. Parallel solution of triangular systems of equations. Parallel Comput., 6:109—114, 1988.
S. C. Eisenstat, M. T. Heath, C. S. Henkel, and C. H. Romine. Modified cyclic algorithms for solving triangular systems on distributed-memory multiprocessors. SIAM J. Sci. Stat. Comput., 9:589—600, 1988.
D. Wang. Solving Triangular Systems on Distributed-Memory Multiprocessors, M.Sc. thesis, Dept. of Mathematics and Statistics, University of Guelph, Guelph, Ontario, Canada, 1993.
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Wang, D., Chu, E. Minimizing Communication Penalty of Triangular Solvers by Runtime Mesh Configuration and Workload Redistribution. The Journal of Supercomputing 14, 77–95 (1999). https://doi.org/10.1023/A:1008151330872
Issue Date:
DOI: https://doi.org/10.1023/A:1008151330872