Still, Qi does not expect to challenge the electronic transistor with his optical analogue, which consumes a lot more power and runs much more slowly. “We want to complement the Intel transistor,” he says. “We don’t want to replace it.” He hopes to find a foothold in niche markets, such as equipment for scrambling cable channels and military technologies that could benefit from light’s imperviousness to an electromagnetic attack.
Routers that guide information through the Internet could also be amenable to optical transistors and switches. At present, these stopping points in the network convert optical signals travelling through fiber-optic cables into electrical signals; these are then processed, converted back to light and sent on their way. A router in which one beam of light pushes another in the appropriate direction — with no conversions involved — could in principle be faster and consume less energy.
A popular candidate for such switches are quantum dots, small semiconductor crystals that behave like atoms. In one particularly sensitive quantum-dot switch, a beam of light is first guided along a material dotted with holes, called a photonic crystal. The light can pass through a quantum dot placed in its path without changing course. But if a pulse of light is sent in just ahead of that beam, it can induce an interaction between the dot and the crystal that scatters the beam and sends it on a different path.
Reported in May 2012 by Edo Waks of the Joint Quantum Institute at the University of Maryland in College Park and colleagues, it switches when struck by a pulse of 140 photons. In principle, that is a small enough amount of energy to rival conventional routers.
But the switch still faces a practical obstacle common to all of these emerging optical technologies. The lasers that supply the devices with light consume considerable energy, offsetting any savings. “Right now,” says Waks, “the overhead is what’s killing us.”