NSF Awards First IT Research Grants
October 21, 2000The National Science Foundation has announced the first round of awards in the ITR (Information Technology Research) initiative. Funded at $90 million for FY 2000, the initiative is expected to grow in future years. Of greatest interest to the SIAM community is the advanced computational science component, which drew about a fifth of the 1400 proposals submitted and received about the same proportion of the funding.
"Interdisciplinary" was a key word in the first competition, and the same will be true for the second round, for which preproposals are due at NSF by the end of November. Two of the largest computational science proposals funded (awards fall into two categories: above and below $500,000) are for "truly interdisciplinary collaborations," says Charles Koelbel of NSF's Directorate for Computer & Information Science & Engineering.
The largest (at more than $11million over five years), based at the University of Florida and titled "Towards Petascale Virtual-Data Grids," brings together physicists and computer scientists, in approximately equal numbers, to address the problem of moving huge quantities of data, virtualizing the data in the process. "They set out to change the way scientists handle data," Koelbel says. "I believe they're going to do it."
The other very large model project is that of computational geometer Herbert Edelsbrunner of Duke University. Titled "Computational Geometry for Structural Biology and Bioinformatics," the project again is interdisciplinary, aiming to advance both IT and the biological sciences. Despite the well-recognized importance of the geometric shapes of molecules in determining their function, geometric methods are relatively unknown in computational biology, the researchers wrote in their abstract. The project, awarded more than $7 million over five years, will focus on difficult issues in the application of geometric computing to biology.
Sixty-two large projects were funded, at an average of $1 million per year for three to five years, along with 148 smaller projects ($500,000 or less for up to three years). Although the name of the main computational science component has been changed for FY 2001---to Applications in Science and Engineering---Koelbel says that the goals and emphases are much the same as for the first round. The point, he says, is still to make an impact on both sides-both in IT and in the scientific discipline.
A final point: That the computational science component ended up with a respectable 20% of the ITR funding was a result of the numbers of proposals submitted. "Being proactive, submitting proposals, is the way to grow the program," Koelbel says, concluding on an optimistic note: There's good reason to think that the role of computational science will improve in the coming years of the initiative.
Information about the projects funded, and about the submission of proposals for the next round of awards, can be found at http://www.itr.nsf.gov/.
