SCIENCE
National Science Board Approves Funds for Petascale Computing Systems
Track 1 and Track 2 systems boost computing power for researchers: Yesterday the National Science Board (NSB) approved a resolution authorizing the National Science Foundation (NSF) to fund the acquisition and deployment of the world's most powerful "leadership-class" supercomputer, proposed in response to NSF's "Track 1" supercomputing solicitation. This "petascale" system is expected to be able to make arithmetic calculations at a sustained rate in excess of a sizzling 1,000-trillion operations per second (a "petaflop" per second) to help investigators solve some of the world's most challenging science and engineering research problems.
A second resolution, for a "Track 2" system, authorized funding for a smaller, but still extremely powerful system expected to bridge the gap between current high-performance computers (HPC) and even more advanced petascale systems under development.
Both awards must still go through administrative and financial processing by NSF before they become final. That process will likely be completed this fall.
"Working at the frontiers of knowledge is increasing the demand for powerful cyberinfrastructure," said NSF Deputy Director Kathie L. Olsen. "The Track 1 system will provide U.S. scientists and engineers access to unprecedented petascale computing resources that will allow them to ask and answer complex questions we haven't even dreamed of."
NSF's Office of Cyberinfrastructure (OCI) supports programs to create and operate several levels of advanced, high-performance computing systems that will allow science and engineering researchers to ask and answer ever larger and more complex questions. The NSB actions today approved funding for "Track 1" and "Track 2" awards to enable petascale science and engineering through the deployment and support of the most capable world-class combination of assets available to the academic community.
Track 1: A Leadership-Class System
In the first award, the University of Illinois at Urbana-Champaign (UIUC) will receive $208 million over 4.5 years to acquire and make available a petascale computer it calls "Blue Waters," which is 500 times more powerful than today's typical supercomputers. The system is expected to go online in 2011.
Under the direction of Dr. Thomas Dunning, Blue Waters will reside at UIUC where it will be operated by the National Center for Supercomputing Applications and its academic and industry partners in the Great Lakes Consortium for Petascale Computation. The system may be used, for example, to study complex processes like the interaction of the Sun's coronal mass ejections with the Earth's magnetosphere and ionosphere; the formation and evolution of galaxies in the early universe; understanding the chains of reactions that occur with living cells; and the design of novel materials.
The project will also impact the preparation of a new generation of students with the skills required to make effective use of high-performance computing in academia and industry. UIUC and its partners, including the Southeast Universities Research Association, the Great Lakes Consortium and the Shodor Education Foundation, will integrate petascale computing into pre-college and college education and develop a new virtual school of computational science and engineering that will have an impact on a national scale..
Track 2: Mid-Range High-Performance Computing Towards the Petascale
The second award will fund the deployment and operation of an extremely powerful supercomputer at the University of Tennessee at Knoxville Joint Institute for Computational Science (JICS). The $65 million, 5-year project will include partners at Oak Ridge National Laboratory, the Texas Advanced Computing Center, and the National Center for Atmospheric Research. Under the direction of Dr. Thomas Zacharia, the group will acquire and provide to the research community a system with a peak performance of just under one petaflop that is almost four times the capacity of the current NSF-supported Teragrid.
The TeraGrid is the world's largest, most powerful and comprehensive distributed cyberinfrastructure for open scientific research. It currently supports more than 1,000 projects and over 4,000 researchers geographically spanning the entire United States.
The system, to be allocated under normal TeraGrid policy, will permit researchers to use high-resolution, multiscale/multiphysics simulations for such tasks as studying the properties of proteins at the atomic scale; understanding the complexities of the brain; determining the fundamental properties of elementary particles; modeling natural disasters, and understanding the delicate balance of processes that are responsible for the global climate and its variation over time. The project includes several activities in education and outreach including efforts aimed at broadening the participation of women and minorities in science and engineering.
This award to the University of Tennessee is the second Track 2 system NSF has awarded. Last year, the Foundation made a 5-year, $59 million Track 2 award to the Texas Advanced Computing Center at The University of Texas at Austin and its partners at Arizona State University and Cornell University. NSF's Track 2 initiative is a four-year activity designed to fund the deployment and operation of up to four leading-edge computing systems that will greatly increase the availability of computing resources to U.S. researchers.
To facilitate the use of these increasingly complicated tools by a broad range of researchers, the Track 2 systems will be integrated into a national, distributed cyberinfrastructure known as the TeraGrid. The TeraGrid brings together resources for computation, data analysis and visualization, data management and storage with extensive technical help and consulting resources, on-line and in-person training opportunities, collaboration tools, and web-based entry-points. These simplify the use of advanced digital technology for research, lowering the barrier to their use and greatly increasing their availability.