Energy Needs May Limit Size, Ability Of Quantum Computers

Julio Gea-Banacloche, professor of physics, found that the more accuracy you want, the more energy you need. He reports his findings in a recent issue of Physical Review Letters.

Quantum computing relies on using single atomic particles as units for information storage. Manipulating this information requires pulsed electromagnetic fields—which contain energy.

"The question of where this energy is going to come from and where it is going to go has to be addressed," he said. "Also, with solid-state controllers, such as capacitors, there is a minimum size because they have to be able to hold a certain amount of energy."

Gea-Banacloche wanted to determine whether there was a minimum amount of energy required for such quantum operations and if this has an impact on how well quantum computers perform.

He found that the energy needed to perform a calculation is inversely proportional to the error rate: In other words, more energy means less uncertainty.

"The more precise you want to be, the more energy you need to put into the system," he said.

Gea-Banacloche began to explore different types of proposed quantum computing control systems—including atom-to-atom interactions and electromagnetic fields—and found the same energy requirements in all of them.

"That is when I began to suspect there was some kind of underlying principle operating here," he said. "Ultimately, it all comes from Heisenberg’s Uncertainty Principle." This principle states that the more precisely the position of a particle is determined, the less precisely the momentum is know in this instant, and vice versa.

Quantum computers have a property called decoherence time—a time limit for operating before the system falls apart. Shorter decoherence times require a higher energy expenditure. Gea-Banacloche’s estimate of the minimum power requirement of a system with a ten microsecond decoherence time is 10 megaWatts. Most solid state quantum systems—systems built of atoms in a solid matrix—currently operate with even shorter decoherence times. Atomic trap quantum devices, which use electric and magnetic forces to hold atoms in place, have longer decoherence times.

"It’s not an impossibility, but it’s an indication of how important it is to improve the decoherence times and improve error correction," he said.

Accuracy matters to all computer users—and especially to researchers who rely on them to compute, store and retrieve data. But in the quantum computing world errors will occur unless something is done to prevent them. Quantum computers must constantly scan and correct every bit for errors because of the decoherence property—the tendency for a system to fall apart. If you have a device with a million bits of information, bursts of energy must be sent to all of the bits, all at once, all the time, to correct for errors. Devices with short decoherence times must perform these operations more often. So small margins of error require large amounts of energy.

"If you could come up with a way to tolerate more error, there could be a big payoff," Gea-Banacloche said.