UF Leads the Way

CLAS physicist guides development of ultra-powerful computer data grid--NSF supports the project with $11.9 million grant

AveryComputer technology continues to blossom at amazing speeds. Phrases like "100 trillion operations per second" and "1 million times 1 billion bytes" are not as inconceivable as they once were. The opportunities that such advanced technology bring to research and collaborative projects across the disciplines are phenomenal.

On September 13, when the National Science Foundation announced that the University of Florida would be awarded $11.9 million over the next five years to research and develop the Grid Physics Network (GriPhyN, pronounced "griffin"), it placed the university at the forefront of the development of a computational system that will have resounding effects on research and scholarship on a global scale. The grant is part of a $90 million Informational Technology Research initiative by the NSF to insure that the US is at the forefront of developing fundamental and innovative information technology.

UF, in conjunction with the University of Chicago, will head the research and development of the project, which is a collaboration with 14 other universities throughout the US. Paul Avery, lead scientist and UF professor of physics, remarks on the scope of the project. "The GriPhyN project encompasses more than just Florida. It includes 16 universities and the idea is to put together a computational data grid linking the participating schools and turning them into a large virtual community. They will become a community of researchers with a network of resources that is much more powerful than any individual resource. Eventually this network will grow to include other universities, laboratories, and libraries around the world."

mapGriPhyN will likely become the world's fastest and most powerful computer data grid. In a September 28 article, The New York Times described the potential of the network. "GriPhyN will be far from the simple data delivery system familiar to users of the World Wide Web. It will be the first large-scale data grid, an intelligent network that will deliver not just raw data, but also the power to do challenging computations. To GriPhyN users, thousands of computers and millions of gigabytes of data will look like one single computing engine of unprecedented power."

Unprecedented, data-intensive computational needs, fundamental to both science and commerce in the twenty-first century, drive the project. "By early next decade, we are going to be dealing with databases that will be 10 million times bigger than they are now," Avery notes. "These databases will extend beyond the ability of one site to hold them all. We will need to link them and search them across the internet and this is the kind of project that will allow that to happen."

The Grid will be connected by a system of high speed networking that will enable it to work efficiently. "Imagine that the data is like water," Avery explains. "We have huge lakes of data in various locations. If they were linked by thin straws, then it would be very difficult to transport the water from one place to another, but if you had wide pipes, you could move the water very quickly. You need high speed networking to make the whole project more efficient."

In its early stages, the project aims to benefit four physics experiments that will explore the structure of the universe and the fundamental forces of nature. The scope of GriPhyN, however, will quickly move beyond the realm of physics and become invaluable across the sciences. William G. Luttge, executive director of the McKnight Brain Institute and professor of neuroscience at UF's College of Medicine, has targeted the project as fundamental for research in his field. "One of the hallmarks of the McKnight Brain Institute is our incredible array of ultrahigh technology instrumentation. Yet because of this and our global research partnerships, as we look to the future we are very concerned that we will be literally buried in massive amounts of data. I am convinced that the funding of the GriPhyN project will go a long way toward preparing us to meet this daunting challenge."

Interest in GriPhyN reaches beyond the sciences. Such a network, which will have the capacity to store and move massive amounts of data, is crucial for projects in both the humanities and the social sciences. John Leavey, chair of the English department, has been interested in the development of the Grid network proposal from early on. "The proposed Grid infrastructure would have a profound impact on future possibilities for current programs in the humanities," he explains. "The Grid infrastructure will make collaboration on major topics possible and available without regard to the location of the scholars involved. It will also enable important archive preservation and access to significant media collections."

quoteAs Leavey's remarks indicate, one of the most important contributions that GriPhyN will make to future scholarship and research is that it will allow global communities of scholars to work together on projects and initiatives regardless of where they are located. Collaborative work that involves sharing massive amounts of data, whether that be film archives or DNA sequences, will be possible through this network. "The transmission of knowledge is a crucial element of humanistic research," says Leavey. "The conference and the seminar are typical and significant examples of this exchange. Other versions of collaboration, however, are now developing in various areas of the university and the humanities, and information technology is providing the means to achieve this new collaboration."

Genetics, genomics, and the collection and analysis of satellite data are other examples of collaborative research areas that must share large amounts of data transcontinentally in order to continue to develop in the twenty-first century.

Not only will the Grid network encourage collaborative projects, it will also enable increasingly diverse participation in collaborative efforts. "I am a believer that this really will allow groups that have not historically participated as much as they could because of resources to contribute in a much bigger way," says Avery. A scholar or an independent researcher will not need to find the funds for an international airplane ticket to attend conferences or visit laboratories or libraries in order to share or collect data. "All you will need is a monitor, a terminal," explains Avery. "It will function as a portal into the Grid system. Once you are in, you can use the resources of the system no matter where you are."

The seemingly endless possibilities and complexities of GriPhyN are still difficult to fathom. Yet soon, like the internet ten years ago is today, it very likely will be something that we take for granted and cannot imagine living without. Pencil

--Laura H. Griffis

Return to Index