Cyberinfrastructure will fuel 21st century scientific
discovery, and Grid and supercomputing will play a
central role in that future, National Science
Foundation Director Rita Colwell said in a keynote
address to the Supercomputing 2002 conference in
Baltimore today.
"A robust, flexible and comprehensive
cyberinfrastructure will give us the foundation we
need to make rapid progress in understanding our human
complexities," Colwell said.
The NSF will soon release its final report on
cyberinfrastructure needs. Colwell cited a number of
current Grid and supercomputing projects that could
serve as the basis for that future
cyberinfrastructure.
At the top of that list was the TeraGrid, the
NSF-funded effort that will create "the most advanced
computing facility available to scientists for all
types of research in the United States," Colwell
said.
"It is a step toward the vision of a
cyberinfrastructure that will give a broad range of
researchers access to high-performance computing,
high-bandwidth networks, very large data stores, and
sophisticated tools for knowledge discovery," she
said.
The Grid Physics Network, or GriPhyN, is another
example of the "nascent cyberinfrastrucrure," Colwell
said, enabling creation of petascale virtual data
Grids.
The NEESGrid, the Network for Earthquake Engineeering
Simulation, is also a "21st century model for
collaboration, a laboratory without walls or clocks,"
she said. The project will help understanding of the
destructiveness of earthquakes and enable development
of construction methods and materials to minimize
damage and loss of life. Data from the EarthScope
distributed observatory, fed into NEES, will create
"an enitrely new plane of discovery."
Colwell also discussed examples from life sciences.
The Human Genome project required 500 million trillion
sequence comparisons, she said, citing Celera
Genomics, a feat that would have been impossible
without an advanced computational infrastructure.
Protein folding used to take 20 months to simulate,
she said; it can now be done in a day at 1 trillion
operations per second.
But there is much left to be done, Colwell told the
conference's 5,200 attendees from 36 countries. There
are 10-100 million species on earth; 1.7 million are
known, and only 50,000 have been described in any
detail, she said.
"We will need the power of supercomputing and the
integration and insight that a comprehensive
cyberinfrastructure provides to untangle these complex
interactions," she said.
A robust cyberinfrastructure would help identify
emerging problems like the global fresh water crisis,
emerging infectious diseases, global climate changes
and homeland security, Colwell said.
"The greatest question of our times may be how we can
avoid the pitfalls and still grasp the opportunities
that science aand technology hold," she said.
NSF Holds Grid Computing Session
The NSF, which was instrumental in the development of
the Internet, has been one of Grid computing's
strongest backers, viewing the technology as the
future of the Internet.
The agency, along with the standards-setting Global
Grid Forum, held a brainstorming session at the
conference on the National Middleware Initiative
(nsf-middleware.org), which recently released the
second version of its package of free middleware and
is in the process of drafting a strategic plan on the
future direction of the program, said program director
Alan Blatecky. Blatecky said NMI will issue a new
middleware release every six months, with the next
release due in April 2003.