Researchers are experimenting with computers that could solve massively complex problems and surpass today's most powerful supercomputers. And those futuristic computers are based on DNA.

To create one, scientists synthesize DNA molecules and present them with a mathematical problem. While today's most complex computers use figures and formulas, a DNA computer's input, output, and software are made up of DNA molecules that store and process encoded information in living organisms.

"To some extent, a DNA computer is a chemical and biological counterpart to today's electronic computers," says Ravinderjit Braich, a postdoctoral student at the University of Southern California's Laboratory for Molecular Science. To understand how a DNA computer works, think of a mathematical problem as a padlock and potential solutions as a ring of keys. While electronic computers try only one at a time, a DNA computer could try millions of keys simultaneously.

The DNA computer was conceived in 1994 by USC professor Leonard Adleman, who founded the school's Laboratory for Molecular Science the following year. NASA's Jet Propulsion Laboratory and its Ames Research Center have provided grants for Adleman's research. The research reached a milestone recently when Adleman used a DNA computer to solve a complex problem with 1 million possible answers--only simple problems had been tried before. The molecular size of DNA computers means many can be combined in a tiny amount of space, Braich says. "DNA has such a high information density that you can record the entire Library of Congress and encode it into DNA that weighs less than 1 gram," he says.

Braich has high hopes for biological and chemical computing. In the future, DNA computers using simultaneous chemical reactions as massively parallel processors could someday be used to study genes and power robotic devices, he says. For now, Adleman says, the DNA computer isn't sufficiently developed.