D. W. Walker
ABSTRACT
Recently a number of advanced architecture machines have become commercially available. These new machines promise better cost-performance then traditional computers, and some of them have the potential of competing with current supercomputers, such as the Cray X/MP, in terms of maximum performance. This paper describes an on-going project to evaluate a broad range of advanced architecture computers using a number of complete scientific application programs. The computers to be evaluated include (1) distributed-memory machines such as the NCUBE, INTEL and Caltech/JPL hypercubes, and the MEIKO computing surface, (2) shared-memory, bus architecture machines such as the Sequent Balance and the Alliant, (3) very long instruction word machines such as the Multiflow Trace 7/200 computer, (4) “traditional” supercomputers such as the Cray X/MP and Cray-2, and (5) SIMD machines such as the Connection Machine. Currently 11 application codes from a number of scientific disciplines have been selected, although it is not intended to run all codes on all machines. Results are presented for two of the codes (QCD and missile tracking), and future work is proposed.
- Baillie 88.Clive F. Baillie, Thomas Gottschalk, and Adam Kolawa, "A Comparison of Multiple-Target Tracking Performance of Different Concurrent Computers", this Volume, 1988.Google Scholar
- Bower 87.J.M. Bower, M. E. Nelson, M. A. Wilson, G. C. Fox, and W. Furmanski, "Piriform (Olfactory) Cortex Model on the Hypercube", Caltech Report CSP 404, 1987.Google Scholar
- Cabibbo 82.N. Cabibbo and E. Marinari,"A New Method of Updating SU(N) Matrices in Computer Simulations of Gauge Theories", Phys. Let~. 119B:387, 1982.Google ScholarCross Ref
- Clayton 87.R.W. Clayton, B. H. Hager, and T. Tanimoto, "Applic,,tions of Concurrent Processors in Geophysics", Proceedings of fhe Second International Conference o~ 5upercompuling, May 3- 8, 1987, Santa Clara, California, pub. International Supercomputln$ Inc., St. Petersburg, Florida, 1987.Google Scholar
- Ding 88.Hong-Qiang Ding and William A. Goddard, "Polymer Simulation on the Hypercube", this Volume, 1988.Google Scholar
- Flower 87.J.W. Flower, "Lattice Gauge Theory on a Parallel Computer", Ph.D. Thesis, California Institute of Technology, CMtech report CsP 411, 1987.Google Scholar
- Fox 87.Geoffrey C. Fox, W. Furmanski, "Hypercube Communication for the Neural Network Algorithms'', Caltech P, eport CSp 405, 1987.Google Scholar
- Fox 88a.Geoffrey C. Fox, Mark A. Johnson, Gregory A. Lyzenga, Steve W. Otto, 3ohn K. Salmon, and David W. Walker, "Solving Problems on Concurrent Processors", pub. Prentice-Hall, Englewood Cliffs, New :Iersey, 1988. Google ScholarDigital Library
- Fox 88b.Geoffrey C. Fox and David W. Walker, "A Portable Programming Environment for MultiprocessoIs", to be picsented at the 12th IMACS WoIld Congress on Scientific Computation, July 18-22, 1988, Paris, France, 1988.Google Scholar
- Gorham 88.Peter W. Gorham, Thomas A. Prince and Stuart Anderson, "Hypercube Data Analysis in Astronomy: Optical Interferometry and Millisecond Pulsar Searches", this Volume, 1988.Google Scholar
- Gottschalk 87.T.D. Gottschalk, "CALTRAX: the Tracking Program for Simulation-87", Caltech report C3P 478, 1987.Google Scholar
- Hipes 87.Paul G. Hipes and Aaron Kuppermann, "Lifetime Analysis of High Energy Resonances in 3-Dimensional Reactive Scattering", Chem. Ph!ls. Leit. 133:1, 1987.Google ScholarCross Ref
- Hsiung 87.Christopher C. Hsiung, "The Myth of Performance for Parallel Machines", unpublished report, 1987.Google Scholar
- Kuppermann 86.Aaron Kuppermann and Paul G. Hipes, "Three Dimensional Quantum Mechanical Reactive Scatteiing Using Symmetrized Hype~spherical Coordinates", }. Chera. Phys. 84:5962, 1986.Google ScholarCross Ref
- Liewer 88.Paulett C. Liewer, R. W. Gould, Geoffrey C. Fox, V. K. Decyk, and J. D. Dawson, "Plasma Simulations on the Mark III Hypercube Computer", this Volume, 1988.Google Scholar
- Martin 87.Joanne L. Martin and Dieter Mueller-Wichards, "Supercomputer Performance Evaluation: Status and Directions", The Journal of Supercompuiing, Vol. 1, May 1987.Google Scholar
- Morison 88.Rodney Morison, "Interactive Performance Display Using the NCUBE Real-Time Graphics System", this Volume, 1988.Google Scholar
- NAS 86."An Agenda for Improved Evaluation of Supercomputez Performance", report to the National Academy of Sciences, pub. National Academy Press, Washington, D.C., 1986.Google Scholar
- Pepin 88.Francois Pepin, K. Chua, and A. Leonard, "Vortex Dynamics on Concurrent Computers", this Volume, 1988.Google Scholar
- Van de Velde 87.E. F. Van de V~ld~, H. B. Keller, "The Design of a Parallel Multigrid Algorithm", Proceealings of the Second International Confe~'ence or~ Supercompuiing, May 3-8, 1987, Santa Clara, California, pub. International Supercomputing Inc., St. Petersburg, Florida, 1987.Google Scholar
- Walker 88.David W. Walker, T. Aldcroft, A. Cisneros, G. C. Fox, W. Furmanski, "LU Decomposition of Banded Matrices and the Solution of Linear Systems on Hypercubes", this Volume, 1988. Google ScholarDigital Library
Index Terms
- Performance evaluation of scientific programs on advanced architecture computers
Recommendations
Modelling the Runtime of Scientific Programs on Parallel Computers
ICPP '00: Proceedings of the 2000 International Workshop on Parallel ProcessingMessage-passing programs for parallel machines with a distributed address space using communication operations of portable communication libraries such as PVM and MPI guarantee portability for most of the parallel machines available. However, due to ...
Performance evaluation of the SX-6 vector architecture for scientific computations: Research Articles
The growing gap between sustained and peak performance for scientific applications is a well-known problem in high-performance computing. The recent development of parallel vector systems offers the potential to reduce this gap for many computational ...
Performance evaluation of supercomputers using HPCC and IMB benchmarks
IPDPS'06: Proceedings of the 20th international conference on Parallel and distributed processingThe HPC Challenge (HPCC) benchmark suite and the Intel MPI Benchmark (IMB) are used to compare and evaluate the combined performance of processor, memory subsystem and interconnect fabric of five leading supercomputers - SGI Altix BX2, Cray X1, Cray ...
Comments