Winter 1999
|
|
Working nonstop with hand calculators, it would take every man, woman, and child in the United States more than 125 years to equal what the Department of Energy (DOE) and Intel Corporation's new computer can do in just one second.
"Hardware is developing at a staggering rate," says Jan Cuny, a University of Oregon computer scientist. "By the year 2002, there may be machines hundreds of times this powerful."
What kinds of applications are researchers finding for all this new computing power?
The DOE's new ultrafast computer is being used to develop simulations that test the safety and reliability of the U.S. nuclear stockpile without underground testing--at great savings to both the treasury and the environment.
Cuny and other UO researchers are playing a role in this monumental effort. "Our work on the DOE project is focused on creating the layer of software that links scientists to the vast amounts of computational power their simulations require," she explains.
Cuny has a similar focus in other work with colleagues at the UO, where she is a member of the Computational Science Institute. This cross-disciplinary group of more than 20 researchers applies computational power to problems in many academic disciplines.
"Researchers in many fields such as biology and geology are now collecting huge amounts of data," she says. "They want to sift through this information for subtle patterns. They want to integrate it with other data using computationally intense methods. I work with these researchers to create a user-friendly package that gives them access to the computer technology they need."
For hundreds of years scientists have followed one of two approaches--experimental or theoretical--to do their work. But now Cuny and other researchers are helping computational science emerge as a third fundamental approach. The tool employs high-speed network technology, large-volume data servers, high-performance graphics workstations and, at its core, powerful parallel supercomputers such as those at the UO, which are capable of six to eight billion calculations per second.
"Computational science is an extremely powerful and rapidly evolving tool of increasing power and profound importance. It is changing the way scientists think about their work and redefining the arena of possibilities in which science is conducted. "
Advanced computing will be a cornerstone of science as it will be practiced in the twenty-first century. That is why it is a cornerstone of the Brain, Biology, and Machine Initiative (see "Taking the Initiative"). As UO biologists, physicists, and neuroscientists expand the frontiers of knowledge, many will leverage their research with the power of advanced computational science.
"The Brain, Biology, and Machine Initiative will vastly expand our capabilities. It will provide us with much needed technical support such as programmers, operators and faculty members. It will enable us to broaden our use of computational science," Cuny explains.
This advanced research environment will also create great opportunities for students. More and more science-related jobs will require training in computational science, she predicts. "And our students will be ready for those jobs. An important element of computational science is that it is collaborative--we teach our students how to work in multidisciplinary teams. This is a skill highly valued by employers."
As a discipline, computer science has been growing for decades, establishing a strong theoretical, experimental, and technological foundation. "Now," Cuny says, "we are ready for the next stage, applying the results of those efforts to multidisciplinary explorations. Investment in the Brain, Biology, and Machine Initiative is how we at the UO are building toward that remarkable future."