This issue...

  Brieflies

  View from the Inside

   Humble Weed Becomes Genome Model

  Microscopic Shock Waves

  Working Science

  People

  Site Seeing

  E-mail Reminder































This issue...

  Brieflies

  View from the Inside

   Humble Weed Becomes Genome Model

  Microscopic Shock Waves

  Working Science

  People

  Site Seeing

  E-mail Reminder































This issue...

  Brieflies

  View from the Inside

   Humble Weed Becomes Genome Model

  Microscopic Shock Waves

  Working Science

  People

  Site Seeing

  E-mail Reminder

Working Science

1998--It Was a Good Year for Supercomputing!

At SuperComputing '98 (SC98), sponsored by the annual conference on high performance networking and computing, those who have pushed the envelope in a constantly stretching discipline were honored by their peers for their contributions. SC98 is sponsored by International Electrical and Electronics Engineers (IEEE) Computer Society and Association for Computing Machinery (ACM) SIGARCH.

Teams supported by divisions within the Office of Science received the majority of the awards--both 1998 Gordon Bell Prizes, as well as the three finalist positions for that award. The 1998 Fernbach Award was given to a mathematician at the National Energy Research Scientific Computing Center (NERSC), and several other recognitions went to Office of Science-supported teams from several national laboratories.

And the winners are...

The 1998 Sidney Fernbach Award, given each year to one person who has made "an outstanding contribution in the application of high performance computers using innovative approaches," was given to Phillip Colella, a mathematician and leader of the Applied Numerical Algorithms Group at NERSC at Lawrence Berkeley National Laboratory (LBNL).

Colella, who has been a staff member of NERSC and the Computing Sciences Directorate at the LBNL since 1996, is being recognized "For fundamental contributions in the development of software methodologies used to solve numerical partial differential equations, and their application to substantially expand our understanding of shock physics and other fluid dynamics problems," according to a letter from Doris Carver, president of the IEEE Computer Society.

T3E Cray ComputerThe Gordon Bell Prize, awarded annually to honor the best achievement in high performance computing, went to a team of scientists at Oak Ridge National Laboratory working with NERSC. The team, which also includes collaborators at the Pittsburgh Supercomputing Center and the University of Bristol (UK), performed a 1,024-atom first-principles simulation of metallic magnetism in iron that ran at 657 Gigaflops (657 billion calculations per second) on a 1024-processor Cray/SGI T3E supercomputer. However, the group later gained access to a machine on Cray's manufacturing center floor and set an all-time record of 1.02 Teraflops -- a trillion calculations per second!

QCDSP ComputerA second Gordon Bell Prize recognizes scientists who achieve the best price/performance level on a computer system. The winning team in this category is a collaboration of universities and DOE national laboratories led by Columbia University and involving the Office of Science's Brookhaven National Laboratory and Fermi National Accelerator Laboratory. The winning machine, costing only $1.8 million, is a multi-purpose, non-commercial supercomputer with a top operating speed of 600 billion calculations per second. This was a benchmark run on the 25 Gigaflop Quantum Chromo Dynamics Supercomputer Prototype funded by the Office of Science's division of High Energy Physics. The machine will be used for physics research.

ASCI RedThe two finalists for the Gordon Bell Prize were the Sandia National Laboratory, University of California, and Intel Corporation collaboration, using the ASCI Red supercomputer (right) at Sandia with more than 4600 dual-processor Pentium Pro nodes, Avalonwhich performed at 605 gigaflops to calculate electronic structures; and Los Alamos National Laboratory's Avalon computer (left), which cost only $150,000 and performed 20 billion operations per second.

Best of Show: "Innovative Wide-area Applications on the GUSTO Grid Testbed," demonstrated by a team representing the Mathematics and Computer Science Division at Argonne National Laboratory; University of Southern California Information Sciences Institute; California Institute of Technology; State University of New York (SUNY) Stony Brook; and Advanced Photon Source, Argonne National Laboratory. Work on the real-time analysis of Advanced Photon Source data is supported by the Office of Science's Mathematical, Information, and Computational Sciences (MICS) division's Grand Challenge Program.

Most Stellar: "Metacomputing the Einstein Theory of Spacetime: Colliding Black Holes and Neutron Stars Across the Atlantic Ocean," demonstrated by team members from Argonne National Laboratory; National Center for Supercomputing Applications (NCSA); Washington University; Max Planck Institut fuer Gravitationsphysik; Konrad Zuse Institut; and DFN Verein.

Best Industrial Collaboration: "Industrial Mold Filling Simulation Using an Internationally Distributed Software Component Architecture," demonstrated by a team representing Indiana University - Bloomington; Industrial Materials Institute, NRC, Quebec; Los Alamos National Laboratory; Argonne National Laboratory; and Centre de Recherche en Calcul Applique, Montreal.

Best Overall Paper: An Out-of-Core Implementation of the Massively Parallel Multireference Configuration Interaction Program, by Holger Dachsel, Jarek Nieplocha, and Robert Harrison from the Pacific Northwest National Laboratory, describes a novel parallelization approach they developed to solve the largest multireference configuration interaction (MRCI) problem ever attempted. MRCI calculations produce highly accurate electronic structure wavefunctions, by including extensive electron correlation in the solution of the Schroedinger equation. Using an out-of-core approach, shared memory programming model, improved data compression algorithms, and dynamic load balancing with a time-sorted task list, they were able to solve a problem six times larger than previously reported. This work was supported by the Office of Science's MICS division under the Phase II Grand Challenge Program and the DOE 2000 ACTS Project and also by the Office of Biological and Environmental Research through the use of the computing facilities in the Environmental Molecular Sciences Laboratory.

For more information about the SC98 awards, see the SC98 website.

Comments
Contents
Search
previous    next
     
   previous     next


• Energy Science News • Energy Science News • Energy Science News • Energy Science News •
www.pnl.gov/energyscience/