Caltech study suggests 10,000–20,000 qubits could enable practical cryptography-breaking quantum computers

Caltech researchers say quantum computers capable of breaking modern cryptography may require fewer qubits than previously estimated, citing new work on error correction for neutral-atom quantum systems. The findings suggest that machines able to run Shor’s algorithm—an approach that could undermine widely used public-key cryptography—may be closer than many earlier projections indicated.

Neutral-atom approach and Shor’s algorithm

The study, published Monday, was developed by Caltech in collaboration with Pasadena-based Oratomic, a quantum computing startup founded by Caltech researchers. The work focuses on a neutral-atom setup in which individual atoms are trapped and controlled with lasers to function as qubits.

The researchers say the system could support a fault-tolerant quantum computer capable of running Shor’s algorithm, which can be used to derive private keys from public keys used in Bitcoin’s elliptic-curve cryptography. The study estimates that this could be achieved with as few as 10,000 reconfigurable atomic qubits.

From physical to logical qubits: error-correction overhead

Oratomic co-founder and CEO Dolev Bluvstein, a visiting associate in physics at Caltech, said the pace of progress is increasing pressure to move toward quantum-resistant cryptography. He also pointed to how earlier qubit estimates have shifted as error correction improves.

According to the article, many current error-correction approaches require about 1,000 physical qubits to produce one reliable, logical qubit, the unit used to perform calculations in a fault-tolerant system. That overhead has contributed to earlier projections that practical systems could require million-qubit scales, slowing timelines for machines that could threaten RSA and elliptic-curve cryptography used by Bitcoin and Ethereum.

System size and lab progress

Bluvstein noted that lab systems are already approaching—and in some cases exceeding—6,000 physical qubits. He said this suggests the cryptography risk may arrive sooner than experts previously expected, as improvements in controllability and system size can reduce the qubit requirements for running Shor’s algorithm.

He also referenced a September milestone from Caltech researchers: a neutral-atom quantum computer operating with 6,100 qubits, achieving 99.98% accuracy and 13-second coherence times. The article describes this as progress toward error-corrected quantum machines and notes it renewed concerns about future threats to Bitcoin from Shor’s algorithm.

Implications for cryptography transition

The potential threat has prompted governments and technology firms to begin migrating to post-quantum cryptography, or encryption designed to withstand quantum attacks. Researchers cited in the article caution, however, that major engineering challenges remain, including scaling quantum systems while maintaining extremely low error rates.

Bluvstein said that while 10,000 physical qubits could be reached within a year, that is not the same as building a system capable of breaking cryptography. He described the engineering task as complex and non-trivial, emphasizing that constructing such machines involves far more than assembling components.

Even so, he said a practical quantum computer could emerge before the end of the decade.

Broader urgency and beyond blockchain

The article also notes that on Tuesday, Google researchers reported new findings suggesting future quantum computers could break elliptic-curve cryptography with fewer resources than previously thought, adding urgency to calls for a transition to post-quantum cryptography.

While the cryptocurrency industry has increasingly focused on quantum risk, Bluvstein said the concern extends beyond blockchain networks. He cited the broader dependence of modern systems on cryptography, including the internet of things devices, internet communication, routers, and satellites, describing the issue as spanning global digital infrastructure.