clock menu more-arrow no yes

New Accuracy Leap Could Speed Arrival of Superfast Quantum Computers

Quantum qubits, now with fewer errors!

Credit: Paul Henderson-Kelly / UNSW Australia

Research teams at the University of New South Wales claim to have developed two new types of “quantum bits” that process data with nearly 100 percent accuracy, addressing one of the key stumbling blocks in the promising realm of quantum computers.

These machines tap into the weird world of quantum mechanics to herald vast leaps in the capabilities of computer processing, as Re/code explored in a recent story.

Traditional computers work with binary bits of information, 1s and 0s. But quantum computers rely on the odd physics that kick in at the atomic level, where objects can occupy more than one space at a time (superposition) and link across vast distances (entanglement).

A quantum computer’s quantum bits, or qubits, can be 1s and 0s simultaneously, and perform numerous operations on the same data at once.

But there are numerous difficulties, including an issue known as coherence, which refers to the challenge of keeping qubits in their fragile quantum states long enough to perform computations and accurately read the results. Even small error rates quickly add up to big ones across hundreds or thousands or millions of computations.

“For quantum computing to become a reality we need to operate the bits with very low error rates,” said Andrew Dzurak, director of the Australian National Fabrication Facility at UNSW, in a statement.

The two teams at UNSW jointly published their findings on Sunday in the journal Nature Nanotechnology.

Dzurak’s team figured out a way to develop a so called “artificial atom” qubit with a device similar to a transistor used in standard laptops and smart phones.

The group led by Associate Professor Andrea Morello, with the school of electrical engineering and telecommunications, focused on pushing what’s possible with the “natural phosphorous” atom qubit. They claim to have achieved a new record for coherence time, specifically 35 seconds.

In both cases, a critical step was placing the qubits inside a thin layer of purified silicon, known as silicon 28, specially designed to eliminate the magnetic noise that can otherwise disturb the quantum bit.

“Half a minute is an eternity in the quantum world,” Morello said in a statement. “Preserving a ‘quantum superposition’ for such a long time, and inside what is basically a modified version of a normal transistor, is something that almost nobody believed possible until today.”

Quantum computing company D-Wave Systems took a different path around the coherence issue, using what’s known as the adiabatic approach that appears to be “relatively resistant to outside interference,” but also only works well on certain types of problems. (There’s been no shortage of controversy over the company’s claims.)

To achieve breakthrough speeds, quantum computers have to operate with numerous qubits in parallel. D-Wave’s forthcoming computer, for instance, boasts 1,152. So the next step for the researchers in Australia is to begin putting together pairs of their new quantum bits, while preserving these levels of accuracy.

Learn more in the video below:

This article originally appeared on Recode.net.

Sign up for the newsletter The Weeds

Understand how policy impacts people. Delivered Fridays.