Design issues for high performance simulation

Abstract One of the key issues in designing new simulation models for parallel execution, or in the migration of existing models to parallel platforms, is the mapping of the application architecture to the parallel system architecture. In this mapping process, we can easily loose track of the inherent locality present in the different architecture layers. In this paper, we present an overview of these issues and examine, by means of several case-studies, the consequences of the design and implementation choices for the various mapping processes. We will show that the potential for high performance simulation comes from a holistic approach, taking into account all aspects from the application to the underlying hardware.

[1]  Peter M. A. Sloot,et al.  Automatic Performance Estimation of SPMD Programs on MPP , 1994, EUROSIM.

[2]  J. Machta,et al.  The parallel complexity of growth models , 1994 .

[3]  Geoffrey C. Fox.  General techniques and regular problems , 1988 .

[4]  Miron Livny,et al.  Condor-a hunter of idle workstations , 1988, [1988] Proceedings. The 8th International Conference on Distributed.

[5]  Message P Forum,et al.  MPI: A Message-Passing Interface Standard , 1994 .

[6]  Anoop Gupta,et al.  Complete computer system simulation: the SimOS approach , 1995, IEEE Parallel Distributed Technol. Syst. Appl..

[7]  Peter M. A. Sloot,et al.  Comparison of vector and parallel implementations of the simulated annealing algorithm , 1995, Future generations computer systems.

[8]  Peter M. A. Sloot,et al.  The CAMAS workbench , 1993 .

[9]  François E. Cellier,et al.  Continuous system modeling , 1991 .

[10]  Horst D. Simon,et al.  Partitioning of unstructured problems for parallel processing , 1991 .

[11]  Peter M. A. Sloot,et al.  Breaking the Curse of Dynamics by Task Migration: Pilot Experiments in the Polder Metacomputer , 1997, PVM/MPI.

[12]  Geoffrey C. Fox,et al.  Parallel Computing Works , 1994 .

[13]  Peter M. A. Sloot,et al.  A dynamic load balancing system for parallel cluster computing , 1996, Future Gener. Comput. Syst..

[14]  Peter M. A. Sloot,et al.  Mermaid: modelling and evaluation research in MIMD architecture design , 1995, HPCN Europe.

[15]  R. Feynman.  Simulating physics with computers , 1999 .

[16]  Peter M. A. Sloot,et al.  Crashworthiness simulation migration to distributed memory and MIMD machines , 1993 .

[17]  Peter M. A. Sloot,et al.  Modelling and Simulation , 1988, Systems Analysis and Simulation 1988, I: Theory and Foundations. Proceedings of the International Symposium held in Berlin (GDR), September 12–16, 1988.

[18]  Nigel P. Topham,et al.  Performance of the decoupled ACRI-1 architecture: the perfect club , 1995, HPCN Europe.

[19]  Jonathan Machta.  The computational complexity of pattern formation , 1993 .

[20]  Witold Dzwinel,et al.  Pattern recognition via molecular dynamics on vector supercomputers and networked workstations , 1995, HPCN Europe.

[21]  Andy D. Pimentel,et al.  Distributed simulation of multicomputer architectures with Mermaid , 1997 .

[22]  Vaidy S. Sunderam,et al.  PVM: A Framework for Parallel Distributed Computing , 1990, Concurr. Pract. Exp..

[23]  Andy D. Pimentel,et al.  Abstract workload modelling in computer architecture simulation , 1997 .

[24]  Peter M. A. Sloot,et al.  The CAMAS workbench: Computer Aided Migration of Applications System , 1994, Future Gener. Comput. Syst..

[25]  Andy D. Pimentel,et al.  An architecture workbench for multicomputers , 1997, Proceedings 11th International Parallel Processing Symposium.

[26]  L. Sander.  Fractal growth processes , 1986, Nature.