Microsoft’s Majorana 2 Claims a 1,000× Leap in Qubit Stability — and Pulls Its Quantum Roadmap Forward to 2029

At its Build 2026 conference, Microsoft unveiled Majorana 2, the successor to the topological quantum chip it introduced in February 2025, claiming qubits that are roughly 1,000 times more reliable than the previous generation. On the strength of that progress, the company says it has cut its roadmap in half and now expects to build a scalable, fault-tolerant quantum computer by 2029.

The headline metric is the “parity lifetime” — a marker of how long a topological qubit holds its quantum information, and the key measure of its quality. Microsoft says Majorana 2 extends that lifetime to about 20 seconds on average, with some qubits surviving as long as a full minute — a dramatic jump from the milliseconds-to-seconds range of earlier devices. “We’re 1,000 times better,” said Chetan Nayak, the Microsoft technical fellow who leads its quantum hardware effort.

The gain comes from a re-engineered materials stack. Majorana 2 replaces the aluminum superconductor of its predecessor with lead, and rebuilds the semiconductor active region around a combination of indium arsenide and indium arsenide antimonide. Microsoft says the new materials better shield the fragile qubits from cosmic and environmental disturbances that knock them out of their delicate quantum state — and notes the materials work was accelerated using its own Discovery agentic AI system, tying the announcement back to the AI theme that ran through the rest of Build.

The whole project rests on Microsoft’s long-standing and contrarian bet: that topological qubits, which encode information in the collective state of the system rather than a single fragile particle, are inherently more stable and error-resistant than the superconducting or trapped-ion qubits favored by rivals like IBM and Google. If that holds, a topological machine could need far fewer physical qubits per error-corrected logical qubit — a shortcut to the fault tolerance every quantum roadmap is chasing.

That “if” is doing a lot of work, and the reaction split accordingly. Constellation Research analyst Holger Mueller called Majorana 2 “a massive step forward,” saying a 20-second parity lifetime is “basically unheard of.” But Microsoft’s topological program carries a long history of contested and even retracted results, and much of the scientific community remains unconvinced; coverage from outlets including Science News and Digitimes noted that critics are still skeptical, particularly because the strongest claims have yet to clear independent peer review.

For now, Majorana 2 is best read as a genuine engineering milestone wrapped in an aggressive timeline. A thousand-fold reliability gain and second-scale coherence would be real progress on the hardest problem in quantum computing — but the 2029 target asks the field to trust a route that has burned it before. The next year of peer-reviewed data, not the keynote, will decide which it is.