Microsoft’s Majorana 2 chip, built on a new lead-based superconductor design, reports a characteristic parity switching time of roughly 20 seconds in an InAs-Pb tetron device, with some instances stretching to minute-scale durations. The company presents these results as a step toward topological qubits that could outperform conventional quantum hardware. But physicists outside Microsoft are pushing back hard, arguing that the quantum effect at the center of the company’s roadmap has never been conclusively demonstrated in any published experiment, and that the diagnostic tools used to support earlier claims contain fundamental flaws.
Why the lead-based tetron result has triggered fresh debate
The tension is straightforward: Microsoft has staked its quantum computing strategy on topological protection, a property that would make qubits far more resistant to errors than competing designs from Google or IBM. The Majorana 2 chip is an update to the earlier Majorana 1 effort, which itself drew sustained criticism from condensed-matter physicists who found the supporting data unconvincing. Swapping aluminum for lead in the superconductor-semiconductor hybrid wires is meant to improve the energy gap that shields the qubit from noise. The new preprint describing parity measurements in an InAs-Pb tetron device puts a concrete number on the table: a characteristic parity switching time of roughly 20 seconds, with some traces lasting longer.
That number matters because longer parity lifetimes suggest better protection of the quantum information stored in the device. If the protection comes from a genuine topological gap, it would validate years of investment. If the protection comes from something else, such as trivial Andreev bound states mimicking the signature of Majorana zero modes, then the entire interpretive framework collapses. The core question is whether the diagnostic protocol Microsoft used to certify its earlier devices can actually distinguish between these two scenarios.
A formal technical critique posted on arXiv directly challenges the topological gap protocol that Microsoft Quantum published in Physical Review B 107, 245423 in 2023. That protocol was designed to confirm topological superconductivity in InAs-Al hybrid devices. The critique argues the protocol is flawed in ways that allow non-topological states to pass the same tests. If the measurement standard itself cannot reliably separate real topological gaps from trivial ones, then the 20-second parity lifetime reported for the new lead-based device does not automatically prove what Microsoft says it proves.
Two preprint critiques and the missing independent replication
The skepticism is not casual or anonymous. It is documented in at least two formal arXiv submissions targeting the specific papers Microsoft used to build its case. The first critique addresses the 2023 Physical Review B paper and its gap protocol. The second targets the 2025 Nature paper reporting interferometric single-shot parity measurements in InAs-Al hybrid devices, which Microsoft used to support the original Majorana 1 claims. That second critique, posted as a comment on the Nature work, questions whether the interferometric data actually establishes the required zero-energy states.
The analytical thread connecting both critiques is consistent. If the gap protocol does not uniquely identify topological superconductivity, then passing it does not confirm Majorana zero modes. And if the interferometric parity measurements in the earlier aluminum-based devices are open to alternative explanations, the same interpretive uncertainty carries forward to the new lead-based hardware. The material upgrade from aluminum to lead may improve device performance in measurable ways, but it does not resolve the underlying question of whether the observed states are topological in nature.
No independent laboratory has publicly replicated the parity lifetime measurements on either the aluminum or lead versions of Microsoft’s tetron devices. The full raw interferometric datasets and minute-scale parity traces from the InAs-Pb device have not been released for outside analysis. Microsoft Quantum has not published a direct rebuttal to the specific technical arguments raised in either preprint critique. This absence of public engagement with the criticism leaves the scientific community in a position where the company’s claims rest on internal measurements evaluated against a diagnostic standard that outside researchers have formally questioned.
Unresolved gaps in the Majorana 2 evidence chain
The central unresolved question is whether an alternative analysis of the parity-switching data, one that applies a stricter gap metric designed to exclude trivial Andreev bound states, would yield a topological gap below the threshold Microsoft claimed in its 2023 Physical Review B paper. If it would, the reported 20-second parity lifetime could reflect conventional superconducting protection rather than the topological variety. No outside group has yet performed that re-analysis on the new lead-based device data, because the data has not been made available.
A second open question involves the relationship between material improvements and topological claims. Lead has a larger superconducting gap than aluminum, which means devices built with it should show longer parity lifetimes regardless of whether the protection is topological. Distinguishing a bigger conventional gap from a genuine topological gap requires exactly the kind of protocol validation that critics say has not been established. The upgrade to lead may have improved the device without settling the physics.
Researchers tracking this dispute should watch for three developments. First, any independent replication of the interferometric parity measurements, ideally in a laboratory with no stake in the outcome. Second, the release of full raw datasets from the InAs-Al and InAs-Pb devices, allowing outside theorists to apply alternative analysis pipelines and stricter gap criteria. Third, a detailed public response from Microsoft Quantum that addresses the specific mathematical and experimental objections raised in the arXiv critiques, rather than treating them as background noise.
The stakes go well beyond a single company’s roadmap. Topological quantum computing has long been framed as a possible shortcut around the daunting overhead of error correction in conventional superconducting qubits. If Microsoft’s devices truly host well-isolated Majorana zero modes with minute-scale parity lifetimes, that would mark a qualitative leap for the field. If, instead, the devices are sophisticated but ultimately conventional superconducting systems whose behavior can be explained without invoking topology, then the community will need to recalibrate expectations and investment priorities.
Outside commentary has underscored this tension. A recent news analysis notes that Microsoft’s new lead-based results arrive in a landscape already shaped by years of disputed Majorana claims and retracted papers. Another Nature report highlights how the lack of independent confirmation has fueled calls for higher standards of evidence before declaring topological quantum computing “solved.” Together, these perspectives frame the Majorana 2 announcement not as a final breakthrough, but as the latest move in a long-running scientific argument that is still far from resolved.
For now, the evidence chain for Majorana 2 contains visible gaps. The reported 20-second parity lifetime in an InAs-Pb tetron is an impressive device metric, but its interpretation depends on a diagnostic framework that is under active, formal challenge. Without independent replication, open data, and a robust response to those challenges, the community is left with an ambiguous picture: a promising experimental platform whose most dramatic claims remain unproven. How Microsoft and its critics handle the next round of analysis and disclosure will determine whether the Majorana 2 chip is remembered as the dawn of topological quantum computing or as another case study in the difficulty of turning subtle quantum signatures into solid technological foundations.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
