NEW ORLEANS — NO longer a luxury for the privileged few, supercomputers are becoming standard tools for solving scientific problems, a panel of scientists says. The machines, so named because they can perform calculations up to a thousand times faster than conventional desktop computers, are playing a growing role in air pollution research, agriculture, biochemistry, and medicine, according to researchers at a meeting of the American Association for the Advancement of Science.
A group at the Carnegie Mellon University in Pittsburgh, for example, is using a supercomputer simulation to chip away at Los Angeles air pollution. The simulation, which runs on a Cray Y-MP/832, allows scientists to monitor the flow of nearly 50 kinds of chemicals over the city and explore the impact of temperature, sunlight, and weather on smog and ground-level ozone.
With the supercomputer, scientists can perform experiments that would be impossible to conduct in a lab or in the city, says Gregory McRae, a scientist who worked on the Carnegie Mellon University (CMU) project.
``You can take away half the cars in Los Angeles inside the computer,'' and the computer can calculate what the change would be on the city's pollution, Dr. McRae explains. Two hundred hours of supercomputer time are needed to simulate a single day of smog. ``To simulate processes of this scale would have taken several years on conventional computers,'' he says.
Supercomputers have become an important tool, in part, as a result of a 1985 National Science Foundation program that established five supercomputer centers across the nation. Nevertheless, ``there is a very serious problem in the US with the lack of this kind of computer power,'' McRae says.
In the case of Los Angeles, the CMU group discovered that the government's approach to controlling ozone was flawed because the plan concentrated only on reducing airborne hydrocarbons, ignoring nitrogen oxides. As a result, laws have been changed. ``Research with supercomputers can save us hundreds of millions of dollars,'' by telling policymakers the most crucial pollutants to regulate, McRae says.