Parallel and Distributed Processing Symposium, International
Download PDF

Abstract


   Workshop Description

Grids

Grids, decentralized collections of heterogeneous resources supplemented with usability middleware, are emerging as an attractive platform for achieving high application performance. This workshop provides a forum to report performance characteristics of grids and grid applications. It covers benchmarks for measuring grid performance and partitioning methods for adaptation of applications to the grid platform.

Benchmarks

Grid measurement technology is necessary for evaluating performance of grid hardware and middleware. Specifically, grid benchmarks provide a basis for measuring the grid performance characteristics such as computational efficiency, communication speed, scheduler efficiency, programmability, and scalability. The issues of repeatability of measurements and resource utilization tracking make grid benchmarking more complicated than traditional supercomputer benchmarking.

Partitioning

Partitioning of applications for efficient use of grids is an emerging area of research. Applications that can exploit grids range from parameter-space studies, Monte Carlo simulations, to graph-based applications and multi-physics coupled models. Currently, grid core middleware such as Globus does not provide a service for partitioning. On the other hand, existing partitioning tools are not well suitable for grids. Partitioning applications for computational grids aims at achieving high performance of existing applications on grids.

Topics

  • Measurement Tools for Computational Grids
  • Grid Practice
  • Grid Benchmarking
  • Partitioning for Computational Grids

   Workshop Organizers

  1. Eric Aubanel, University of New Brunswick
  2. Virendra C. Bhavsar, University of New Brunswick
  3. Michael Frumkin, NASA Ames Allan Snavely, San Diego Supercomputing Center
  4. Rob F. Van Der Wijngaart, NASA Ames

   Program Committee

  1. Akshai Aggarwal, University of Windsor, Windsor
  2. Rupak Biswas, NASA Ames, Moffett Field
  3. Henri Casanova, San Diego Supercomputing Center
  4. Bharat S. Chaudhari, The International Institute of Information Technology Nikos P. Chrisochoides, College of William and Mary, Williamsburg Thierry Coupez, CEMEF - groupe CIM, Ecole des Mines de Paris
  5. Brian J. d'Auriol, The University of Texas at El Paso
  6. Weichang Du, University of New Brunswick
  7. Wolfgang Gentzsch, Sun Microsystems
  8. Valentina Huber, Research Centre Juelich
  9. George Karypis, University of Minnesota, Minneapolis
  10. Thuy T. Le, San Jose State University Robert Preis, University of Paderborn Ruth Shaw, University of New Brunswick
  11. Chris Walshaw, University of Greenwich
  12. Rich Wolski, UC Santa Barbara Laurence Tianruo Yang, St. Francis Xavier University