Abstract
Multi-tier platforms are hierarchically organized multicomputers with multiprocessor nodes. Compared with previous-generation single-tier systems based on uniprocessor nodes, they present a more complex array of performance tradeoffs. We describe performance programming techniques targeted to finite difference methods running on two large scale multi-tier computers manufactured by IBM: NPACI’s Blue Horizon and ASCI Blue-Pacific Combined Technology Refresh. Our techniques resulted in performance improvements ranging from 10% to 17% over a traditional single-tier SPMD implementation.
Chapter PDF
Similar content being viewed by others
Keywords
- Shared Memory
- Message Passing Interface
- Lawrence Livermore National Laboratory
- Annular Region
- Memory Hierarchy
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.
References
Fink, S. J.: Hierarchical Programming for Block-Structured Scientific Calculations. Doctor dissertation, Dept. of Computer Science and Engineering, Univ. of Calif., San Diego (1998)
Baden, S. B. and Fink, S. J.: Communication Overlap in Multi-tier Parallel Algorithms. In Proc. SC’ 98, IEEE Computer Society Press (1998)
Fink, S. J. and Baden, S. B. Runtime Support for Multi-tier Programming of Block-Structured Applications on SMP Clusters. In: Ishikawa, Y., Oldehoeft, R, Reynders, J. V. W., and Tholburn, M. (eds.), Scientific Computing in Object-Oriented Parallel Environments. Lecture Notes in Computer Sci., Vol. 1343. Springer-Verlag, New York (1997) pp. 1–8
Fink, S. J. and Baden, S. B. A Programming Methodology for Dual-tier Multicomputers. IEEE Trans. on Software Eng., 26(3), March 2000, pp. 212–26
Baden, S. B. and Fink, S. J., and Kohn, S. R. Efficient Run-Time Support for Irregular Block-Structured Applications. J. Parallel Distrib. Comput., Vol 50, 1998, pp. 61–82
MPI Forum: The Message Passing Interface (MPI) Standard. http://www-unix.mcs.anl.gov/mpi/index.html, 1995
IEEE: IEEE Guide to the POSIX Open System Environment. New York, NY, 1995
Gropp, W. W. and Lusk, E. L. A Taxonomy of Programming Models for Symmetric Multiprocessors and SMP Clusters. In Giloi, W. K. and Jahnichen, S., and Shriver, B. D. (eds.), Programming Models for Massively Parallel Computers. IEEE Computer Society Press, 1995, pp. 2–7
Eigenmann, R., Hoeflinger, J., Jaxson, G., and Padua, D. Cedar Fortran and its Compiler, CONPAR 90-VAPP IV, Joint Int. Conf. on Vector and Parallel Proc., 1990, pp. 288–299
Crandall, P. E., Sumithasri, E. V., Leichtl, J., and Clement, M. A. A Taxonomy for Dual-Level Parallelism in Cluster Computing, Tech. Rep., Univ. Connecticut, Mansfield, Dept. Computer Science and Engineering, 1998
Murer, S., Feldman, J., Lim, C.-C., and Seidel, M.-M. pSather: Layered Extensions to an Object-Oriented Language for Efficient Parallel Computation, Tech. Rep. TR-93-028, Computer Sci. Div., U.C. Berkeley, Dec. 1993
Lim, B.-H., Heidelberger, P., Pattnaik, P., and Snir, M. Message Proxies for Efficient, Protected Communication on SMP Clusters, in Proc. Third Int’l Symp. on High-Performance Computer Architecture, San Antonio, TX, Feb. 1997, IEEE Computer Society Press, pp. 116–27.
Woodward, P. R.. Perspectives on Supercomputing: Three Decades of Change, IEEE Computer, Vol. 29, Oct. 1996, pp. 99–111.
Sawdey, A. C., O’Keefe, M. T., and Jones, W. B. A General Programming Model for Developing Scalable Ocean Circulation Applications, Proc. ECMWF Workshop on the Use of Parallel Processors in Meteorology, Jan. 1997.
Somani, A. K. and Sansano, A. M. Minimizing Overhead in Parallel Algorithms through Overlapping Communication/Computation, Tech. Rep. 97-8, NASA ICASE, Langley, VA., Feb. 1997
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2001 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Baden, S.B., Shalit, D. (2001). Performance Tradeoffs in Multi-tier Formulation of a Finite Difference Method. In: Alexandrov, V.N., Dongarra, J.J., Juliano, B.A., Renner, R.S., Tan, C.J.K. (eds) Computational Science — ICCS 2001. ICCS 2001. Lecture Notes in Computer Science, vol 2073. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-45545-0_90
Download citation
DOI: https://doi.org/10.1007/3-540-45545-0_90
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-42232-7
Online ISBN: 978-3-540-45545-5
eBook Packages: Springer Book Archive