Computers sought to perform petaflops

/ Source: The Associated Press

Supercomputers that perform trillions of operations every second are helping scientists probe life's deepest complexities. But it's still not enough.

On the horizon now are computers of greater than a "petaflop," or 1 quadrillion computations per second. The National Science Foundation is seeking to fund such a computer, one that can sustain one to two "petaflops" with occasional peaks of about 10 times that amount.

Even one petaflop would be about three times faster than the peak of the world's current top supercomputer, the 367-teraflop BlueGene/L made by IBM Corp. for the Department of Energy's Lawrence Livermore nuclear lab. IBM also has been working to surpass 1 petaflop.

"We're talking about a half a million to a million processors that will be in the machine," said Thom Dunning, director of the National Center for Supercomputing Applications at the University of Illinois, which hopes to win NSF funding for a petaflop system. "Your PC has one processor in it."

With petaflops of computing power, scientists say they could simulate problems and test solutions in such fields as engineering, biology, medicine and economics with far greater detail and speed — in seconds rather than days.

"We can solve problems that we simply couldn't solve before," said Edward Seidel, director of the Center for Computational Technology at Louisiana State University and an astrophysicist interested in black holes and gravitational waves in outer space.

Reaching these heights will require overcoming huge obstacles. Half a million microprocessors produce a lot of heat, requiring vast amounts of electricity for cooling. Software applications must be rewritten to work on the new computers. And the machines have to last.

"When you're computing at this scale it becomes an overriding issue," Dunning said. "If you don't play close attention to that, you can get a machine that is really powerful but will only stay up for an hour. That clearly wouldn't lead to many of the scientific breakthroughs we'd be looking for."