Maryland today | Research: Quantum Computers May Be Better Than …

Pobody is nervous, not even the indifferent and calculating bits that are the basis of computers. But for the first time, a team including researchers at the University of Maryland has shown that an assembly of quantum computing parts can be better than the more fragile individual parts used to make it.

In an article published today in the journal Nature, the team which includes Christopher Monroe, a member of the Joint Quantum Institute and professor of physics at College Park, as well as other scientists and colleagues at UMD at Duke University, explained how they took this decisive step towards reliable and practical quantum computers.

In their experiment, the researchers combined multiple qubits – the quantum version of the bits that encode information in typical computers as zeros and ones – to work together as a single unit. This “logical qubit” is based on a quantum error correction code that can detect and correct an error that occurs in any of the 13 individual qubits that make up the logical qubit “team”. In addition, the logical design of the qubit is fault tolerant, i.e. able to contain errors to minimize their negative effects.

The demonstration reinforces the great promise of quantum computers, which are theoretically capable of operations beyond the reach of standard or “classical” computers, in part because qubits are much more flexible than regular bits and are not limited. to zero or one. But quantum errors have long held back efforts to expand these futuristic machines to higher power levels; Unlike transistors that encode information in normal computer chips, a qubit is susceptible to errors due to tiny environmental disturbances like vibration or a change in temperature that brings it out of its quantum state.

But a group of qubits working as a team can help work around those limitations, said Monroe, who is also the co-founder and chief scientist of IonQ, a college-park quantum company based in part on the technology he has. developed as a UMD researcher.

“Qubits composed of identical atomic ions are natively very clean on their own,” he said. “However, at some point when many qubits and operations are required, errors need to be reduced further, and it is easier to add more qubits and code information differently. The beauty of correction codes is A mistake for atomic ions is that they can be very efficient and can be flexibly activated through software commands.

This is the first time that a logical qubit has proven to be more reliable than the most error-prone step required to achieve it. The experiment showed that the team could confirm that they had correctly created the logical qubit in a desired quantum state 99.4% of the time, compared to the approximate 98.9% success rate of the six quantum processes (called quantum operations) that they used to do so. .

It may not seem like much of a difference, but it is a crucial step in the quest to build much bigger quantum computers. If the six quantum operations were assembly line workers, each focusing on one task, the combined worker error rate would cause the line to produce only useful products 93.6% of the time, much lower at the 99.4% efficiency rate when “workers” collaborate to minimize the risk of quantum errors compounding and ruining the bottom line.

While it might seem unprofitable to use so many individual qubits and steps just to make something work as a single qubit, the unique computational capabilities of quantum computers could make logical qubits a small price to pay. If quantum computers can be made trustworthy, they will be powerful computational devices that are expected to revolutionize industries such as healthcare, security, and finance.

The results were obtained using Monroe’s ion trap system at UMD, which uses up to 32 individual charged atoms – ions – which are cooled with lasers and suspended over electrodes on a chip. The ions can then be used as qubits through further laser manipulation.

“We have 32 laser beams,” Monroe said. “And the atoms are like ducks lined up; each with its own fully controllable laser beam. I think the atoms form a linear string and we pluck it like a guitar string. We pick it up with lasers that we activate and deactivate in a programmable way. And that’s the computer; it is our central processing unit.

By successfully creating a fault-tolerant logic qubit with this system, researchers have shown that careful and creative designs have the potential to free quantum computing from the constraint of the inevitable errors of the current state of the art.

“What’s amazing about fault tolerance is that it’s a recipe for taking small, unreliable parts and turning them into a very reliable device,” said Kenneth Brown, professor of electrical and computer engineering at Duke and co-author of the article. “And fault tolerant quantum error correction will allow us to make very reliable quantum computers from faulty quantum pieces.”

In addition to Monroe and Brown, the article’s co-authors are Laird Egan, JQI graduate student; JQI researcher Marko Cetina; Andrew Risinger, Daiwei Zhu and Debopriyo Biswas, graduate students of JQI; Dripto M. Debroy, graduate student in physics at Duke University; postdoctoral fellows from Duke Crystal Noel and Michael Newman; and Muyuan Li, graduate student from Georgia Institute of Technology.