The President's Council of Advisors on Science and Technology's (PCAST's) recently released report suggests that pure supercomputing speed is not the best indicator of technological prowess and that the United States' research funds would be better spent developing effective software and applications for supercomputing technologies. Gaining the top spot on the annual ranking of supercomputers is "an arms race that is very expensive and may not be a good use of funds," says PCAST's David E. Shaw.
PCAST's report examined nonclassified federal funding for computing research and found that about half of the $4.3 billion per year budget goes toward technology development in support of research at various science agencies. University of Washington computer scientist Edward Lazowska says that less funding is spent on long-range computing research than the overall numbers indicate, and notes that "computer science is crucially about discovery as well."
The report says computing research should prioritize developing new techniques for exploring large-scale data sets and algorithms for machine learning, as well as research on privacy and cybersecurity. Lazowska says solving those types of problems requires developing software and computers that can mine vast data sets to uncover patterns and insights.
From The New York Times
View Full Article
Abstracts Copyright © 2010 Information Inc., Bethesda, Maryland, USA
No entries found