The dream of creating game-changing quantum computers — supermachines that encode information in single atoms rather than conventional bits — has been hampered by the formidable challenge known as quantum error correction.

In a paper published Monday in Nature, Harvard researchers demonstrated a new system capable of detecting and removing errors below a key performance threshold, potentially providing a workable solution to the problem.

“For the first time, we combined all essential elements for a scalable, error-corrected quantum computation in an integrated architecture,” said Mikhail Lukin, co-director of the Quantum Science and Engineering Initiative, Joshua and Beth Friedman University Professor, and senior author of the new paper. “These experiments — by several measures the most advanced that have been done on any quantum platform to date — create the scientific foundation for practical large-scale quantum computation.”

In the new paper, the team demonstrated a “fault tolerant” system using 448 atomic quantum bits manipulated with an intricate sequence of techniques to detect and correct errors.

The key mechanisms include physical entanglement, logical entanglement, logical magic, and entropy removal. For example, the system employs the trick of “quantum teleportation” — transferring the quantum state of one particle to another elsewhere without physical contact.

“There are still a lot of technical challenges remaining to get to very large-scale computer with millions of qubits, but this is the first time we have an architecture that is conceptually scalable,” said lead author Dolev Bluvstein, Ph.D. ’25, who did the research during his graduate studies at Harvard and is now an assistant professor at Caltech. “It’s going to take a lot of effort and technical development, but it’s becoming clear that we can build fault-tolerant quantum computers.”

The Harvard-led collaboration included researchers from MIT and was jointly headed by Lukin; Markus Greiner, George Vasmer Leverett Professor of Physics; and Vladan Vuletić, Lester Wolfe Professor of Physics at MIT. The team conducts research in collaboration with QuEra Computing, a startup company spun out from Harvard-MIT labs, the Joint Quantum Institute at University of Maryland, and the National Institute of Standards and Technology.

The new paper represents an important advance in a three-decade pursuit of quantum error correction.

“In the end, physics is an experimental science. By realizing and testing these fundamental ideas in a lab, you really start seeing light at the end of the tunnel.”

Mikhail Lukin

Conventional computers encode information in a binary code of zeros and ones. Quantum computers store information in subatomic particles whose counterintuitive properties of quantum physics can achieve far more processing power.

In a conventional computer, the most basic unit of information is a “bit” (short for binary digit); in quantum systems, the basic unit is a “qubit” (or quantum bit).

In conventional computers, doubling the number of bits doubles the processing power; in quantum systems, adding qubits exponentially increases the power because of a phenomenon called quantum entanglement.

In theory, a system of 300 quantum bits can store more information than the number of particles in the known universe.

With such vast power, quantum computers have the potential to deliver breakthroughs in fields such as drug discovery, cryptography, machine learning, artificial intelligence, finance, and material design.

But there are hurdles to realizing that revolutionary potential. Chief among them is the error rate. Qubits are inherently susceptible to slipping out of their quantum states and losing their encoded information, making error correction a core prerequisite to achieving large quantum machines.

In the new paper, the team combined various methods to create complex circuits with dozens of error correction layers. The system suppresses errors below a critical threshold — the point where adding qubits further reduces errors rather than increasing them.

“There have been many important theoretical proposals for how you should implement error correction,” said Alexandra Geim, one of the lead authors on the new paper and a Ph.D. student in physics in the Kenneth C. Griffin Graduate School of Arts and Sciences. “In this paper, we really focused on understanding what are the core mechanisms for enabling scalable, deep-circuit computation. By understanding that, you can essentially remove things that you don’t need, reduce your overheads, and get to a practical regime much faster.”

Lukin said years of experiments showed how to overcome some technical challenges and avoid others.

“We realize which of these bottlenecks are real and which bottlenecks you just can bypass,” he said. “In the end, physics is an experimental science. By realizing and testing these fundamental ideas in a lab, you really start seeing light at the end of the tunnel.”

Researchers around the world are studying a variety of potential platforms for qubits, including different types of atoms, ions, and superconducting qubits.

The Harvard team specializes in neutral atoms (those with no electrical charge because they have equal numbers of protons and electrons) of the element rubidium. They use lasers to change the configuration of electrons to encode the atoms to become information-carrying qubits.

Hartmut Neven, vice president of engineering at the Google Quantum AI team, said the new paper came amid an “incredibly exciting” race between qubit platforms.

“This work represents a significant advance toward our shared goal of building a large-scale, useful quantum computer,” he said.

In September, the Harvard-MIT-QuEra group published another Nature paper demonstrating a system of more than 3,000 qubits that could operate continuously for more than two hours and overcome another technical hurdle of atom loss.

With recent advances, Lukin believes the core elements for building quantum computers are falling into place.

“This big dream that many of us had for several decades, for the first time, is really in direct sight,” he said.

This research received federal funding from the Defense Advanced Research Project Agency, Department of Energy, Intelligence Advanced Research Projects Activity, Army Research Office, National Science Foundation, and National Defense Science and Engineering Graduate Fellowship program.

Share this article