There's also no agreement about the best way to build a quantum computer. Experiments by Isaac Chuang and Neil Gershenfeld at MIT and David Wineland at NIST use atoms or charged ions in an electromagnetic trap, while IBM is testing superconducting materials that can generate quantum bits. "We're trying to beat our brains to come up with something different than what others are doing," Amer says. Two weeks ago, IBM and Stanford University launched joint research in a related area called spintronics, which aims to design electronics by controlling how electrons spin, instead of how they transmit electrical charges. They plan to research applications for building quantum computers, among other areas.

Another question hanging over the field is whether experiments are progressing quickly enough. Researchers estimate it would take a quantum computer of 100,000 calculating atoms to perform work that's beyond the reach of today's most-powerful computers, such as breaking the 128-bit codes used to encrypt E-mail and other Internet data in a reasonable amount of time. "We're barely getting one or two" calculating atoms, says Stan Williams, director of quantum-science research and a senior fellow at HP. Scientists need to focus on goals less complicated than breaking codes that secure online data, Williams says. "You'd need a quantum computer so large that it would be like trying to build a supercomputer before a transistor radio," says Williams, who's responsible for making sure HP's nanotech and quantum research has a payoff. "You need to take baby steps and bootstrap an industry out of quantum computing," he says. "Where's the low-hanging fruit that will start to get some revenue to pay for more R&D?"

One possibility is a new HP project exploring how small numbers of quantum bits could be sent through fiber-optic cable to offices and homes in wired cities. If they can, HP researchers theorize, the quantum properties of photons could be exploited to model more-efficient outcomes for markets that behave according to game theory, such as financial markets and auctions. "We're trying to find what people would call a killer app," Williams says. "Rather than beat our heads against the wall and bloodying ourselves trying to factor a large number, we're trying to find something people would pay money for with just a few qubits and the basic research on the technology for delivering them."

The price may be unknown--"a nickel a qubit," Williams jokes--but the goal is serious. Companies expect even futuristic researchers to keep an eye on cash flow. "You're almost responsible in a laboratory like HP's for having a business model," Williams says. "We're going off in a different trajectory than other people."

Other technology companies, including Intel, aren't jumping into the high-risk field. Intel's business could be directly threatened by a breakdown in the advances predicted by Moore's Law. But quantum-computing technologies "are far away from anything Intel as a mass-market supplier of complex math functions would be interested in," says Pat Gelsinger, Intel's chief technology officer. The company is funding some university research, but no Intel staffers are researching it directly. "I don't think it matters before 2020," Gelsinger says.

Some in the high-performance-computing user community are skeptical as well. One technologist on Wall Street says a shift as big as the one from electronic to quantum computing--no matter how technically feasible--would require massive amounts of work to fit into the private sector. Every line of companies' software code would have to be ported or rewritten, for example.

It's a dilemma for every emerging technology: When is the right time to give it money and attention? Quantum computing is harder to justify than most, since practical uses seem so far off--and far out. Yet it's those hard-to-imagine uses that also make it so critical to understand. NIST's Williams acknowledges that most companies don't yet see what value comes from tracking quantum computing, quantum information, and quantum cryptography. "At the same time, they may say, 'This is scary,'" he says. "When the boat sails, if they're not on, they may cease to exist."

Illustration by Steve Keller

Continue to the sidebar: Opposites Attract: Quantum Computing's Strange World