COLLEGE OF SCIENCES TO LEAD FEDERAL ENERGY PROJECT
Old Dominion University's College of Sciences will provide project leadership on a U.S. Department of Energy (DOE) cooperative agreement intended to revolutionize the ability of scientists to take advantage of the world's most powerful computers, federal officials announced Aug. 14.
The university will conduct algorithmic research, create software and provide consulting on its use throughout the DOE national laboratory complex, together with five other universities and computational groups at three national laboratories across the country.
The other universities are: Carnegie Mellon University, Pittsburgh; New York University; the University of California, Berkeley; the University of Colorado, Boulder; and the University of Tennessee, Knoxville.
The labs are: Argonne National Laboratory in Illinois; and Lawrence Berkeley and Lawrence Livermore national laboratories, both in California.
The agreement is part of the DOE's new Scientific Discovery through Advanced Computing (SciDAC) program. Fifty-one projects nationwide will receive a total of $57 million this fiscal year to advance fundamental research in several areas related to the department's missions, including climate modeling, fusion energy sciences, chemical sciences, nuclear astrophysics, high-energy physics and high performance computing.
Old Dominion's five-year, nine-institution contract is worth approximately $18 million overall, about $13 million of which will be spent at the DOE laboratories that house the computers on which the software is designed to run.
Old Dominion will receive $1 million for its portion, which consists of technical direction of the overall effort and algorithmic work on large-scale systems of equations. An algorithm is a predetermined set of instructions for solving a given problem in a finite number of steps.
David Keyes, Richard F. Barry Professor and chair of the Department of Mathematics and Statistics at Old Dominion, is heading up the overall project and is the principal investigator of the Old Dominion portion of it.
The research group is a prestigious one, Keyes said. It includes two members of the National Academy of Engineering and the current president of the Society for Industrial and Applied Mathematics as co-principal investigators.
It also includes members of three of the best computational science research groups in the world in the Mathematics and Computer Science Division at Argonne, the Computing Sciences Division at Berkeley National Lab and the Center for Advanced Scientific Computing at Lawrence Livermore.
Established as part of DOE's 2001 line-item initiative Scientific Discovery through Advanced Computing, the Old Dominion-led "Integrated Software Infrastructure Center (ISIC) is one of only seven created around the country.
The center focuses on developing and implementing algorithms and supporting scientific investigations performed by the DOE.
Large-scale simulations of importance to the Department of Energy often involve the solution of partial differential equations involving sophisticated mathematical models that require terascale computers - those capable of performing more than a trillion calculations per second.
The DOE will be able to apply this technology to a number of areas, including fusion, accelerator design, global climate change and reactive chemistry, Keyes said.
The algorithms created as part of this project will aim to reduce "computational bottlenecks" by one to three orders of magnitude on terascale computers, enabling scientific simulation on a scale heretofore impossible, he added.
Such algorithms have already revolutionized the DOE's Science-based Stockpile Stewardship program, which is intended to support the Comprehensive Test Ban Treaty by making it unnecessary to perform physical tests of nuclear weapons, since the many of the salient results of those tests can now be obtained by simulation.
"In many other areas of science, physical experimentation is impossible, such as with cosmology; dangerous, as with manipulating the climate; or simply expensive, as with fusion reactor design," Keyes explained.
It is hoped that large-scale simulation will give scientists insight and confirmation of existing theories in such areas, without benefit of experimental verification. Such simulations of computer codes may be checked against experiment in a variety of well understood laboratory contexts to validate them.
Extraordinary advances in computing technology in the past decade have set the stage for a major advance in scientific computing, according to the DOE. Within the next five to 10 years, computers another thousand times faster than today's computers will be available. These advances herald a new era in scientific computing - if they can be harnessed with scalable algorithms and software.
"Using such computers, it will be possible to dramatically extend exploration of the fundamental processes of nature, such as the structure of matter from the most elementary particles to the building blocks of life, as well as advance our ability to predict the behavior of a broad range of complex natural and engineered systems, like the earth's climate or an automobile engine," Keyes said.
To exploit this opportunity, these computing advances must be translated into corresponding increases in the performance of the scientific codes used to model physical, chemical and biological systems. This is a daunting problem. Current advances in computing technology are being driven by market forces in the commercial sector, not by scientific computing.
Old Dominion University also has proposed the institution of a Center for Computational Science to coordinate scientific computing projects on and off campus.
Old Dominion has been involved in high-end scientific computing at the national level since 1995, when it became one of 37 universities to participate in the National Science Foundation's Grand Challenges, National Challenges and Multidisciplinary Computing Challenges program with a $1-million grant.
In 1998, Old Dominion became one of 14 universities to participate in the Department of Energy's Accelerated Strategic Computing Initiative academic alliance program, with another million-dollar project.
The university's new ISIC builds on the other investments made in scalable algorithms begun under these and other earlier awards, Keyes said. All, like the present one, were obtained jointly with Alex Pothen, professor of computer sciences at Old Dominion.
To date, six Old Dominion graduate students in computer science and math have been hired by the DOE to aid sharing research performed at Old Dominion with the DOE national laboratory complex. More information is available on the DOE's Office of Science Web page at http://www.energy.gov/HQPress/releases01/augpr/pr01143.htm beginning Tuesday.