This Article 
   
 Share 
   
 Bibliographic References 
   
 Add to: 
 
Digg
Furl
Spurl
Blink
Simpy
Google
Del.icio.us
Y!MyWeb
 
 Search 
   
Relationships Between Efficiency and Execution Time of Full Multigrid Methods on Parallel Computers
June 1997 (vol. 8 no. 6)
pp. 562-573

Abstract—The large number of processing elements in current parallel systems necessitates the development of more comprehensive and realistic tools for the scalability analysis of algorithms on those architectures. This paper presents a simple analytical tool with which to study the scalability of parallel algorithm-architecture combinations. Our practical method studies separately execution time, efficiency, and memory usage in the accuracy-critical scaling model, where the problem size—input data set size—increases with the number of processors, which is the relevant one in many situations. The paper defines quantitative and qualitative measurements of the scalability and derives important relationships between execution time and efficiency. For example, results show that the best way to scale the system (to deteriorate as little as possible the properties of the system) is by maintaining constant execution time. These analytical results are verified with one candidate application for massive parallel computers: the full multigrid method. We study the scalability of a general d-dimensional full multigrid method on an r-dimensional mesh of processors. The analytical expressions are verified through experimental results obtained by implementing the full multigrid method on a Transputer-based machine and on the CRAY T3D.

[1] W.L. Briggs, A Multigrid Tutorial.Philadelphia: SIAM, 1988.
[2] L. Brochard, "Efficiency of Multicolor Domain Decomposition on Distributed Memory Systems," Domain Decomposition Methods, T.F. Chan, R. Glowinski, J. Periaux, and O.B. Widlund, eds., pp. 249-259.Philadelphia: SIAM, 1989.
[3] O.G. Calderón, V.M. Pérez-García, I. Martín, and J.M. Guerra, "Transverse Mode Selection in Single Longitudinal Mode Laser," Physical Rev. A, vol. 53, no. 5, pp. 3,490-3,496, 1996.
[4] G.H. Golub and J.M. Ortega, Scientific Computing: An Introduction with Parallel Computing, Academic Press, San Diego, Calif., 1993.
[5] A. Grama, A. Gupta, and V. Kumar, “Isoefficiency Function: A Scalability Metric for Parallel Algorithms and Architectures,” IEEE Trans. Parallel and Distributed Technology, vol. 1, no. 3, pp. 12-21, 1993.
[6] A. Gupta and V. Kumar, "Performance Properties of Large Scale Parallel Systems," J. Parallel and Distributed Computing, vol. 19, pp. 234-244, 1993. Also available as Technical Report TR 92-32, Dept. of Computer Science, Univ. of Minnesota, Minneapolis, Minn.
[7] A. Grama, V. Kumar, S. Ranka, and V. Singh, "A3: A Simple and Asymptotically Accurate Model for Parallel Computation," Proc. Frontiers '96, pp. 60-69,Annapolis, Md., 1996.
[8] A. Gupta, V. Kumar, and A. Sameh, “Performance and Scalability of Preconditioned Conjugate Gradient Methods on Parallel Computers,” IEEE Trans. Parallel and Distributed Systems, vol. 6, no. 5, pp. 455-469, May 1995.
[9] J. L. Gustafson,“Reevaluating Amdahl's law,”Commun. ACM, vol. 31, no. 5, pp. 532–533, 1988.
[10] K. Hwang, Advanced Computer Architecture: Parallelism, Scalability, Programmability. McGraw-Hill, 1993.
[11] O. Kolp and H. Mierendorff, "Efficient Multigrid Algorithms for Locally Constrained Parallel Systems," Applied Math. and Computation, vol. 19, pp. 169-200, 1986.
[12] V. Kumar, A. Grama, A. Gupta, and G. Karypis, Introduction to Parallel Computing: Design and Analysis of Algorithms. Benjamin Cummings, 1994.
[13] V. Kumar and A. Gupta, "Analyzing Scalability of Parallel Algorithms and Architectures," J. Parallel and Distributed Computing, vol. 22, no. 3, pp. 379-391, 1994. Also available as Technical Report TR 91-18, Dept. of Computer Science Department, Univ. of Minnesota, Minneapolis, Minn.
[14] V. Kumar, A. Smaeh, A. Grama, and G. Karypis, "Architecture, Algorithms and Applications for Future Generation Supercomputers," Proc. Frontiers '96, pp. 346-354,Annapolis, Md., 1996.
[15] J. Linden, G. Lonsdale, H. Ritzdorf, and A. Schüller, "Scalability Aspects of Parallel Multigrid," Future Generation Computer Systems, vol. 10, pp. 429-440, 1994.
[16] I. Martín, V.M. Pérez-García, F. Tirado, and L. Vázquez, "Parallel Multigrid Methods in Nonlinear Optics," SIAM News, vol. 29, no. 3, Apr. 1996.
[17] I. Martín, F. Tirado, and L. Vázquez, "Some Aspects about the Scalability of Scientific Applications on Parallel Computers," Parallel Computing, vol. 22, pp. 1,169-1,195, 1996.
[18] O.A. McBryan,P.O. Frederickson,J. Linden,A. Schuller,K. Solchenbach,K. Stuben,C.A. Thole,, and U. Trottenberg,“Multigrid methods on parallel computers: A survey of recent developments,” Impact Comput. Sci. Eng., vol. 3, pp. 1-75, 1991.
[19] S.F. McCormick, Multigrid Methods.Philadelphia: SIAM, 1987.
[20] J.P. Singh, J.L. Hennessy, and A. Gupta, “Scaling Parallel Programs for Multiprocessors: Methodology and Examples,” Computer, vol. 26, no. 7, pp. 42-50, July 1993.
[21] X.H. Sun, "The Relation of Scalability and Execution Time," Internal Report 95-62, ICASE, NASA Langley Research Center, Hampton, Va., 1995
[22] X.-H. Sun and D. Rover, “Scalability of Parallel Algorithm-Machine Combinations,” IEEE Trans. Parallel and Distributed Systems, vol. 5, no. 6, pp. 599-613, June 1994.
[23] Z. Xu and K. Hwang, “Early Prediction of MPP Performance: SP2, T3D, and Paragon Experiences,” J. Parallel Computing, vol. 22, pp. 917-942, Oct. 1996.

Index Terms:
Parallel computation, scalability of parallel systems, performance metrics, multigrid methods, scientific computing.
Citation:
Ignacio Martín, Francisco Tirado, "Relationships Between Efficiency and Execution Time of Full Multigrid Methods on Parallel Computers," IEEE Transactions on Parallel and Distributed Systems, vol. 8, no. 6, pp. 562-573, June 1997, doi:10.1109/71.595573
Usage of this product signifies your acceptance of the Terms of Use.