Just 10,000 quantum bits might crack internet encryption schemes

Computers based on atoms could demand a fraction of the qubits previously estimated

An grid of multitudes of quantum bits, each associated with a single logical quantum bits, pictured next to a small array of quantum bits with multiple logical quantum bits

Error correction schemes make reliable logical qubits (gold) from error-prone physical qubits (blue). A new scheme produces many more logical qubits per physical qubit (right) than traditional methods (left).

Robert Hurt/IPAC-SELab/Caltech

Powerful quantum computers may be closer than scientists thought.

To unleash the technology’s full power, scientists have long thought that quantum computers with millions of quantum bits, or qubits, would be necessary. But researchers report that quantum computers’ promise might emerge with as few as 10,000 qubits.

One of the key tasks that future quantum computers are expected to excel at is cracking the encryption used to secure communications on the internet. Now, scientists have calculated that a widely used type of encryption called elliptic curve cryptography could be thwarted with a quantum computer with 9,988 qubits — although it would take about 1,000 days to crack.

With around 26,000 qubits, the encryption could be broken in a day, the researchers report in a paper submitted March 30 to arXiv.org. Another prevalent form of encryption, RSA–2048, would require 100,000 qubits and 10 days to break, according to the researchers, from Caltech and quantum computing company Oratomic in Pasadena, Calif.

The calculation suggests that quantum computers could likewise soon contribute to other areas where the machines are expected to have an impact, such as AI, chemistry and materials science.

The new result follows a paper posted on arXiv.org in February, in which researchers from Iceberg Quantum in Sydney calculated that RSA encryption could be defeated in a week with a quantum computer with about 100,000 qubits. The two papers, which have not yet been peer-reviewed, suggest a dramatic decrease from the 20 million qubits thought to be required just a few years ago.

That drop is largely thanks to improvements in quantum error correction, the technique by which faulty quantum bits are made into reliable computational tools. “It has a huge implication on quantum computing,” says physicist Jens Eisert of Freie Universität Berlin, who was not involved with the research. “This work is encouraging [in] that it tells us that this may be more feasible than we think.”

A vibe shift is taking hold in the wake of these and other results. “There’s a new wave of hope that quantum computers can really work, and maybe in the next five to 10 years can really crack our encryption,” says mathematician Jens Niklas Eberhardt of Johannes Gutenberg University Mainz in Germany, who was not involved with the research. “It’s kind of amazing but also terrifying.” That advance could lay bare secret data and threaten the security of cryptocurrencies such as Bitcoin, which relies on elliptic curve cryptography.

The two papers harness advanced types of quantum error correction, called quantum low-density parity check codes. Error correction generally works by combining multiple faulty qubits into one reliable logical qubit. Those logical qubits are then used to perform calculations. Depending on how many qubits go into one error-corrected logical qubit, the technique can demand a massive qubit overhead.

Quantum low-density parity check codes can make logical qubits more efficiently than standard schemes. But they typically require that each qubit is able to interact directly with many other qubits. That’s not easy for many of the most prominent types of quantum computers, which inscribe their qubits on chips on which qubits interface mainly with their nearest neighbors.

Oratomic’s work rests on a type of quantum computer with qubits made of individual atoms. Lasers can move the atoms around, allowing all the qubits to connect to one another. The work suggests that error-corrected quantum computers are a seeming inevitability, says physicist Dolev Bluvstein of Oratomic. “It just looks so close and it seems so attainable that there’s no way humanity is going to stop.”

But putting down a number as a benchmark is just one step. “It doesn’t mean the problem is solved; the fun is only starting,” Eisert says. Fun by a physicist’s definition, anyway: “There’s a lot of details to be worked out.”

And error correction advances are coming rapid-fire. On March 30, researchers with Google Quantum AI posted a paper and blog on the company’s website suggesting that a quantum computer with 500,000 physical qubits running for a few minutes could endanger the security of cryptocurrency.

Altogether, the results have scientists raising alarms: Internet security needs to be updated ASAP. “This is certainly an even stronger impetus for people to upgrade now to quantum-resistant cryptography,” computer scientist Scott Aaronson of the University of Texas at Austin said in an email. “They should really get on it!”