Quantum computer systems promise to achieve speeds and efficiencies unimaginable for even the quickest supercomputers of as we speak. But the know-how hasn’t seen a lot scale-up and commercialization largely as a result of its lack of ability to self-correct. Quantum computer systems, in contrast to classical ones, can’t right errors by copying encoded knowledge again and again. Scientists needed to discover one other means.

Now, a brand new paper in Nature illustrates a Harvard quantum computing platform’s potential to resolve the longstanding downside often called quantum error correction.

Main the Harvard group is quantum optics skilled Mikhail Lukin, the Joshua and Beth Friedman College Professor in physics and co-director of the Harvard Quantum Initiative. The work reported in Nature was a collaboration amongst Harvard, MIT, and Boston-based QuEra Computing. Additionally concerned was the group of Markus Greiner, the George Vasmer Leverett Professor of Physics.

An effort spanning the final a number of years, the Harvard platform is constructed on an array of very chilly, laser-trapped rubidium atoms. Every atom acts as a bit — or a “qubit” because it’s referred to as within the quantum world — which may carry out extraordinarily quick calculations.

The group’s chief innovation is configuring their “impartial atom array” to have the ability to dynamically change its format by transferring and connecting atoms — that is referred to as “entangling” in physics parlance — mid-computation. Operations that entangle pairs of atoms, referred to as two-qubit logic gates, are items of computing energy.

Operating a sophisticated algorithm on a quantum laptop requires many gates. Nonetheless, these gate operations are notoriously error-prone, and a buildup of errors renders the algorithm ineffective.

Within the new paper, the group studies near-flawless efficiency of its two-qubit entangling gates with extraordinarily low error charges. For the primary time, they demonstrated the power to entangle atoms with error charges under 0.5 %. By way of operation high quality, this places their know-how’s efficiency on par with different main kinds of quantum computing platforms, like superconducting qubits and trapped-ion qubits.

Nonetheless, Harvard’s strategy has main benefits over these rivals as a result of its giant system sizes, environment friendly qubit management, and talent to dynamically reconfigure the format of atoms.

“We’ve established that this platform has low sufficient bodily errors that you would be able to truly envision large-scale, error-corrected units based mostly on impartial atoms,” mentioned first writer Simon Evered, a Harvard Griffin Graduate Faculty of Arts and Sciences scholar in Lukin’s group. “Our error charges are low sufficient now that if we had been to group atoms collectively into logical qubits — the place data is saved non-locally among the many constituent atoms — these quantum error-corrected logical qubits may have even decrease errors than the person atoms.”

The Harvard group’s advances are reported in the identical challenge of Nature as different improvements led by former Harvard graduate scholar Jeff Thompson, now at Princeton College, and former Harvard postdoctoral fellow Manuel Endres, now at California Institute of Know-how. Taken collectively, these advances lay the groundwork for quantum error-corrected algorithms and large-scale quantum computing. All of this implies quantum computing on impartial atom arrays is exhibiting the complete breadth of its promise.

“These contributions open the door for very particular alternatives in scalable quantum computing and a really thrilling time for this complete area forward,” Lukin mentioned.

The analysis was supported by the U.S. Division of Power’s Quantum Methods Accelerator Middle; the Middle for Ultracold Atoms; the Nationwide Science Basis; the Military Analysis Workplace Multidisciplinary College Analysis Initiative; and the DARPA Optimization with Noisy Intermediate-Scale Quantum Gadgets program.