Penn State Intercom......January 17, 2002

Global data grid counts
University among partners

By Barbara Kennedy
Eberly College of Science

A consortium of U.S. institutions, including Penn State, has been awarded a $13.65-million grant to create the world's first truly global high-speed data grid for major scientific experiments in physics, astronomy, biology and engineering.

The project, known as the International Virtual Data Grid Laboratory (iVDGL), will seamlessly connect a high-speed worldwide network of powerful computers, initially at 40 locations in the United States, Europe, Australia, Asia and eventually in other regions of the world.

The National Science Foundation (NSF) is funding the U.S. component of the project, which involves 15 universities and four national laboratories. The NSF award is matched by $2 million in university contributions. Partner institutions in other countries are investing more than $20 million to build computational and storage sites as part of the international consortium.

Experiments performed on the grid are expected to influence the future of scientific investigation by providing transparent access to information in a wide range of disciplines, including high-energy and nuclear physics, gravitational-wave research, astronomy, astrophysics, Earth observations and bioinformatics. The network will provide the super-fast speeds required for highly complex and computationally intense areas of scientific research.

University scientists involved in the project include research teams working on two such projects, the Laser Interferometer Gravitational Wave Observatory (LIGO) and the Sloan Digital Sky Survey (SDSS).

"The idea behind the International Virtual Data Grid Laboratory is to make its computing power and worldwide collections of scientific data as easy to use as electricity," explained Lee Samuel Finn, associate professor of physics and astronomy and astrophysics and director of the Center for Gravitational Wave Physics. "Just as you can turn on a light without knowing the source of the electricity, scientists using the grid will be able to tell it to run a certain program on a certain data set without having to think about where the program is run or where the data are located. The grid will automatically and transparently will find the computing resources that are needed to run the requested program and will locate the data that are needed, then bring all the pieces together, do the work and produce the results.

"Eventually, the computational speed of this grid could be measured in petaflops, where one petaflop equals 1,000 trillion calculations per second," said Paul Avery, the project's overall principal investigator and a physics professor at the University of Florida. "The grid will be capable of handling quantities of data measured in petabytes, where one petabyte is 1 million gigabytes, or roughly the amount of data contained in 100,000 personal computer hard drives."

Finn is the principal investigator for a $1.2-million component of the International Virtual Data Grid Laboratory project, which will create a national data-analysis center at the University, including a large "Tier 2" computer supercluster with the advanced Linux operating system. The University supercluster will be available via the International Virtual Data Grid Laboratory to researchers everywhere who are working on LIGO, a U.S. detector that next year is expected to begin collecting the world's first data from gravitational waves generated by the most violent and massive events in the universe, such as the collision of two black holes. The LIGO project is expected to open a new window to the universe by directly detected gravitational waves on Earth.

The Numerically Intensive Computing Group at the Center for Academic Computing will provide infrastructure support and grid expertise for the University supercluster.

Another large scientific project targeted to benefit from the International Virtual Data Grid Laboratory is the Sloan Digital Sky Survey, a large international effort that aims to create a comprehensive digital map of one-quarter of the sky and to measure the distances to a million galaxies and 100,000 quasars.

Along with the University, the U.S. educational institutions participating in the International Virtual Data Grid Laboratory are the University of Florida, the University of Chicago, California Institute of Technology, the University of California at San Diego, Indiana University, Boston University, the University of Wisconsin at Milwaukee, Johns Hopkins University, the University of Wisconsin at Madison, Northwestern University, the University of Southern California, the University of Texas at Brownsville, Hampton University and Salish Kootenai College. Participating U. S. national laboratories include Fermi National Laboratory, Brookhaven National Accelerator Laboratory, Argonne National Accelerator Laboratory and Stanford Linear Accelerator Laboratory. International partners include the U.K. e-Science Programme, the U.K. Particle Physics and Astrophysics Research Council, Italy's INFN institutes, the Australian Partnership for Advanced Computing, the Tokyo Institute of Technology, Tsukuba Advanced Computing Center, and the CERN laboratory in Switzerland.

Barbara Kennedy can be reached at