Supercomputers vs. the Mysteries of Life : Research: Scientists hope the machines will become more powerful in order to mathematically simulate nature.

Supercomputers will evolve into more powerful machines to help mathematically simulate nature and unravel its mysteries, say researchers, who are gathered here this week for an annual supercomputer convention.

“Mother Nature knows all the rules; we’re still trying to learn them,” said Norman Morris, division leader for computing and communications at Los Alamos National Laboratory.

“Even the most powerful machines today are not powerful enough to perform the mathematical experiments that we know how to do,” he said.

It’s a formidable task.

Supercomputing goals are mighty--challenges involving global weather prediction, determining atomic structures, understanding turbulence, mapping the human genome.

Supercomputer forays into these challenges, researchers say, will trigger other breakthroughs in such areas as automobile and aerospace design and the search for oil.

“This concept is calculating complex physical phenomena from first principles . . . understanding what rules a molecule follows,” Morris said.

Supercomputing experts meeting at the Supercomputing ’91 conference will exchange the latest information about research, development and use of high-performance computers.

Raymond Elliott, program manager at the Los Alamos computing and communications division, said more than 3,500 people plan to attend the fourth annual conference.

Supercomputers, generally defined as the most powerful computers made, help researchers gather enormous amounts of data and digest problems.

But supercomputers are expensive and generally have been limited to the realm of national research labs. The Los Alamos lab has six supercomputers; Sandia National Laboratories in Albuquerque has seven.

The labs have focused on defense research, an emphasis derived from big defense spending. But that’s changing.

Defense spending is down, and industry and education need the labs’ supercomputer expertise.

“Getting high-performance computing out into industry is important,” said William Camp, manager of Sandia’s mathematics and computational science department. Partnerships must be formed between the labs and universities and industry, he said.

Edwin Barsis, director of Sandia’s computer sciences and mathematics division, said Sandia has taken a major step in opening up its once-cloaked supercomputer operations.