Microsoft unveiled its Majorana 2 quantum chip at its Build 2026 conference earlier this week, claiming a 1,000-fold improvement in qubit reliability over its predecessor and announcing a revised timeline to achieve a scalable fault-tolerant quantum computer by 2029—cutting the company's previous estimate nearly in half and reigniting a debate that has followed the Majorana program since its inception in Redmond, Washington: does the underlying physics actually work?
What the Majorana 2 Claims to Do
The original Majorana chip, announced in February 2025, could maintain qubit stability for between 1 and 12 milliseconds. Microsoft says Majorana 2 achieves a mean qubit lifetime of 20 seconds, with individual instances lasting as long as one full minute. That improvement matters enormously for practical quantum computing: error correction—the process of keeping qubits in a stable, usable state long enough to run meaningful calculations—requires extended coherence times that earlier generations of hardware simply could not sustain.
Majorana 2 is built on topological qubits, a design approach that encodes quantum information in the collective behavior of exotic particles called Majorana fermions, rather than in individual electrons or photons. The theoretical appeal is robustness: in principle, topological qubits are less vulnerable to environmental noise than conventional designs. Microsoft also replaced aluminum with lead as the chip's superconducting material—a switch the company says contributed significantly to the stability gains observed in internal testing.
AI Designed the Chip in 16 Months
What sets Majorana 2 apart as a hardware story is how it was built. Microsoft says its Discovery agentic AI platform drove the research process, deploying autonomous agent teams that could reason across scientific literature, generate hypotheses, optimize experiments, and validate results—all guided by human researchers rather than replacing them. A development cycle that would historically have taken a decade was compressed to 16 months.
"This is what AI-accelerated science looks like in practice," a Microsoft researcher said at the Build keynote, according to a transcript reviewed by technology media. "We're not just using AI to write code. We're using it to discover materials."
The company says the Power Usage Effectiveness improvements and qubit lifetime data have been independently verified through internal review processes, though the chip has not yet been submitted for third-party experimental replication in peer-reviewed journals.
Physicists Are Not Convinced
The announcement landed with a thud in portions of the academic physics community. A review published in Scientific American noted that outside experts say the topological qubit technology Majorana 2 depends on "doesn't even work and never has"—a reference to persistent doubts about whether Microsoft has ever actually detected the Majorana fermions the architecture requires.
Those doubts have a documented history. In 2018, Microsoft-affiliated researchers published a paper in Nature claiming to have detected Majorana fermions in a semiconductor nanowire. The paper was retracted in 2021 after an independent review found the evidence did not support the conclusion. Critics note that the Majorana 2 announcement, like its predecessor's, relies on internal measurements rather than third-party experimental replication.
"Improved stability numbers are meaningful only if the thing being stabilized is what they say it is," a quantum physicist at the University of Maryland told Science News. "That question remains open."
The Race Is On Either Way
Whether or not Majorana 2's underlying physics satisfies outside reviewers, the announcement lands at a moment of intense geopolitical competition in quantum computing. IBM's roadmap targets large-scale fault-tolerant systems by 2033. Google's quantum team has produced error-correction results using conventional superconducting qubits. China has invested heavily in quantum research, building infrastructure including the world's first offshore wind-powered underwater data center partly optimized for AI and quantum workloads. The National Security Agency and the Department of Defense have both designated quantum computing a strategic national priority.
For Microsoft, the commercial and reputational stakes of Majorana 2 are substantial. If the 2029 timeline holds, the company would leapfrog every competitor by years. If it doesn't—if the topological qubit architecture continues to resist independent verification—the gap between the announcements made in Redmond and the machines that actually run algorithms will be difficult to close without confronting the foundational physics questions that critics say remain unanswered.