Microsoft Majorana 2 Quantum Chip Claims 1000x Reliability Leap - Critics Not Convinced

[Jess30]

Microsoft unveiled Majorana 2 on 2 June at its Build conference in San Francisco, claiming a 1,000-fold improvement in qubit reliability over its predecessor. The new chip uses a topological qubit approach and a materials change from aluminium to lead as the superconductor, which the company says improved a key measure called parity lifetime to a mean of 20 seconds, with some instances reaching a full minute. Most competing quantum approaches measure qubit lifetimes in microseconds, so if the numbers hold up the gap is real.

Microsoft has also cut its projected timeline for a scalable quantum computer from 2033 to 2029, which it credits partly to its agentic AI research platform Microsoft Discovery. The claim is that AI agents accelerated the materials science, fabrication optimisation and measurement automation that made Majorana 2 possible. Microsoft Discovery went generally available on the same day, and a free local app is available via a GitHub Copilot account.

The scepticism from external physicists is significant and should not be glossed over. Scientific American reported that outside experts say the underlying topological qubit approach does not actually work and never has, continuing what they describe as a trend of bold claims followed by limited evidence. Microsoft has been pursuing Majorana qubits for years with mixed peer review, and the company's track record on timeline predictions in this space has been patchy. The 2029 target will be watched very closely.

Introducing Majorana 2news.microsoft.com

[Hollow]

The Scientific American piece is the one to read alongside Microsoft's press release. The gap between what the company is claiming and what independent physicists are saying is large.