Google, Oratomic speed quantum timelines
- Google and Oratomic published research in late March arguing useful quantum machines may arrive sooner, prompting Google and Cloudflare to move quantum-security targets to 2029. - Oratomic said new Caltech-backed work cuts cryptographically relevant hardware estimates to about 10,000 atomic qubits, down from earlier assumptions of roughly 1 million. - Google now says commercially relevant systems could arrive by decade’s end, sharpening pressure to replace today’s encryption. (blog.google)
Quantum computers store information in qubits, which are powerful but fragile; the hard part is keeping them accurate long enough to finish a calculation. (blog.google) That is why researchers talk about error correction: you spread one logical bit of information across many physical qubits, then detect and fix mistakes as they appear. Google says a useful machine will need error-corrected, long-lived logical qubits that can survive about 1 million computational steps with less than one error. (quantumai.google) Google’s recent case for faster timelines starts with Willow, its 105-qubit chip announced on December 9, 2024. The company said Willow reduced errors as systems scaled up and moved Google further along its roadmap toward commercially relevant applications. (blog.google) (quantumai.google) On March 24, 2026, Google said it was expanding beyond superconducting chips into neutral-atom quantum computing, a design that uses individual atoms as qubits. Hartmut Neven wrote that Google is “increasingly confident” commercially relevant superconducting systems will be available by the end of this decade. (blog.google) Oratomic made the sharper numerical claim when it launched on March 31, 2026. The startup said research with the California Institute of Technology showed cryptographically relevant quantum computing may be possible with about 10,000 reconfigurable atomic qubits, far below earlier estimates of around 1 million. (prnewswire.com) Oratomic said those lower hardware needs came from combining neutral-atom systems with more efficient fault-tolerant design and high-rate error-correction codes. The company also said Caltech researcher Manuel Endres has already trapped arrays of 6,000 atomic qubits in the lab. (prnewswire.com) The immediate consequence was not a new computer on the market. It was a shift in security planning: on March 25, Google set a 2029 target for migrating to post-quantum cryptography, citing progress in hardware, error correction, and quantum factoring estimates. (blog.google) Cloudflare followed on April 7 and moved its own full post-quantum security target to 2029. The company said recent advances in quantum hardware and software had accelerated the timeline for a quantum attack. (blog.cloudflare.com) Outside experts are not treating the new estimates as settled. TIME reported the Google and Oratomic papers suggest encryption-breaking machines could arrive sooner, but Princeton professor Jeff Thompson said key assumptions in the Oratomic work remain untested and can look easier “if you just assume better qubits.” (time.com) Oratomic is now trying to turn the research into hardware. On April 28, it announced a partnership with Monarch Quantum aimed at systems with tens of thousands of physical qubits and thousands of error-corrected logical qubits by the end of the decade. (monarchquantum.com) The story here is less that Google and Oratomic proved useful quantum computing has arrived, and more that they moved the planning horizon. Google is talking about commercially relevant systems by 2030, Oratomic is pitching utility-scale machines on the same timetable, and internet companies are already adjusting around that date. (blog.google) (prnewswire.com) (blog.cloudflare.com)
