Supercomputing's big problem: What's after silicon?

In a little booth tucked in the corner of the SC13 supercomputing conference here this week may be the next Intel.

There you'll find Max Shulaker, a Stanford University graduate student, ready to explain carbon nanotube digital circuits to anyone stopping by. Shulaker is part of team of researchers building such circuits, and will be presenting to attendees their research on "the first" computer built using carbon nanotubes, or CNTs.

The Stanford research team has emulated the MIPS instruction set, making it programmable. "Everything we do is silicon compatible," said Shulaker. The result is 10x benefit in a metric that considers performance and energy. In short, it's a big leap over silicon.