Microsoft Claims 1,000x Qubit Reliability Gain With Majorana 2 Chip, Cuts Quantum Timeline to 2029

Microsoft unveiled its second-generation topological quantum chip, Majorana 2, at its Build 2026 conference in San Francisco on June 2, claiming the device is 1,000 times more reliable than its predecessor and announcing a compressed commercial roadmap that now targets a scalable fault-tolerant quantum computer by 2029, cutting the prior timeline roughly in half.

The number that anchors the announcement is a qubit lifetime. According to a Microsoft news feature reviewed from the company's own newsroom, Majorana 2 achieves a mean qubit lifetime of 20 seconds, with some instances holding state for as long as one minute. For context, the aluminum-based Majorana 1 devices it replaces measured lifetimes in the range of 1 to 12 milliseconds, according to a technical analysis published by PostQuantum.com, which reviewed the accompanying preprint.

The architecture decision behind the gain is specific: Microsoft swapped the aluminum superconductor in Majorana 1 for a lead-based one. As reported by Business Standard from Microsoft's announcement, the lead superconductor produces a gap of roughly 1,300 microelectronvolts, compared with about 300 microelectronvolts for aluminum. That wider gap is what Microsoft says gives the qubit more insulation from environmental noise. The company's quantum hardware lead Chetan Nayak, quoted in a SiliconANGLE report on the announcement, put it plainly: the team is "1,000 times better" than a year ago.

The other notable thread in the announcement is how the chip was built. Microsoft says its Discovery platform, an agentic AI system that became generally available the same day, was used to automate the measurement cycles that previously took researchers weeks each, and to synthesize knowledge across the chip's interdisciplinary team spread across multiple countries. That part of the story is less contested than the qubit physics and has more immediate relevance for enterprises evaluating AI in hardware R&D workflows.

The physics, however, remains the harder problem. The Majorana 2 preprint has not been peer-reviewed, and the reaction from the academic community was pointed. As Science News reported, physicist Henry Legg of the University of St Andrews said the new data does "nothing in this preprint resolves the fundamental issues." The core dispute, documented carefully by PostQuantum.com's analysis, is that the paper's measurements are consistent with Majorana zero modes but also consistent with trivial Andreev bound states that can mimic their signatures. The paper presents only Z measurements; demonstrating a full qubit requires both X and Z. Microsoft has said it holds unpublished data on qubit control, but that data is not in the preprint.

This context matters because Microsoft's topological program carries real baggage. A 2018 Nature paper claiming Majorana evidence was retracted in 2021 after investigators found data had been selectively presented. A second paper from the same group was retracted from Nature in 2022. When Majorana 1 arrived in early 2025, the accompanying Nature paper similarly fell short of what the press release claimed, as PostQuantum.com's account documents.

None of that means Majorana 2's materials engineering result is fictitious. Several researchers, including physicist Kartiek Agarwal of Argonne National Laboratory, told Science News the team has made real progress and employed a new nonlocal measurement method that strengthens the topological case. The dispute is over whether progress in materials science constitutes proof of a topological qubit.

For enterprise and infrastructure readers, the practical takeaway is narrow but real. NAND Research, in an analysis of the announcement, assessed that organizations evaluating quantum strategies should treat Microsoft's topological roadmap as a watch item rather than an implementation priority, while noting that the Microsoft Discovery platform's role in the project offers a concrete demonstration of agentic AI applied to hardware manufacturing and scientific research - use cases that are applicable in life sciences and materials science today regardless of how the qubit physics resolves.

The 2029 fault-tolerant target is aggressive. Microsoft says it halved its original timeline on the strength of Majorana 2's results. Whether that confidence survives peer review is the question the field will be watching.

Sources cited:
- Microsoft Newsroom (news.microsoft.com) (https://news.microsoft.com/source/features/innovation/majorana-2-microsoft-discovery-agentic-ai/)
- PostQuantum.com (https://postquantum.com/industry-news/microsoft-majorana-2-analysis/)
- Science News (https://www.sciencenews.org/article/microsoft-quantum-chip-upgrade-majorana)
- Business Standard (https://www.business-standard.com/amp/technology/tech-news/microsoft-majorana-2-quantum-chip-explained-126060300840_1.html)
- SiliconANGLE (https://siliconangle.com/2026/06/02/microsofts-new-majorana-2-quantum-chip-claims-dramatic-breakthrough-qubit-stability/)
- NAND Research (https://nand-research.com/microsofts-majorana-2-topological-quantum-chip/)
- Scientific American (https://www.scientificamerican.com/article/microsofts-upgraded-majorana-quantum-computing-chip-fizzles-with-physicists/)