Microsoft’s Majorana 1 chip represents the company’s boldest claim yet in quantum computing: a processor built on a topological core that, according to Microsoft, stabilizes elusive Majorana particles into qubits resistant to the errors plaguing rival architectures. The announcement leans on two peer-reviewed papers and a string of technical exchanges that together form the most detailed public record of topological qubit progress to date. But a pair of formal critiques posted to arXiv challenge whether the experimental data actually prove what Microsoft says they prove, setting up a scientific dispute with real consequences for the future of quantum hardware.
What is verified so far
The strongest piece of evidence behind the Majorana 1 announcement is a paper published in Nature as volume 638, pages 651 through 655 (2025). That study reports a single-shot interferometric fermion-parity measurement performed in indium arsenide and aluminum (InAs-Al) hybrid devices. Fermion-parity readout is a technical prerequisite for any qubit architecture that relies on Majorana zero modes, because the information stored in a topological qubit is encoded in parity states rather than in the energy levels used by conventional superconducting qubits. The Nature paper includes explicit data and code availability statements, meaning outside researchers can, in principle, attempt to reproduce the analysis.
Before that result, Microsoft Quantum published a 2023 paper in Physical Review B (volume 107, article 245423) describing InAs-Al devices that passed the company’s topological gap protocol. That protocol is Microsoft’s internal benchmark for confirming that a device’s energy gap behaves consistently with theoretical predictions for a topological phase. Together, the two papers form a two-step argument: first, that the devices reach a topological regime; second, that parity can be read out in a single measurement, a necessary condition for building working qubits from Majorana particles.
No independent laboratory has publicly replicated either result. The experimental techniques require highly specialized nanowire fabrication and cryogenic setups, which limits how quickly outside groups can test the claims. Microsoft has not disclosed wafer-level yield statistics or multi-qubit gate fidelity numbers, so the gap between a single-device parity measurement and a functioning multi-qubit processor remains large and publicly unquantified.
What remains uncertain
Two technical comments posted to arXiv target the foundations of Microsoft’s argument. One comment, arXiv preprint 2502.19560, questions the topological gap protocol itself, raising concerns about parameter sensitivity in the 2023 Physical Review B paper. The critique asks whether the protocol can reliably distinguish a genuine topological phase from a trivial phase that mimics the expected signatures under certain parameter choices. If the protocol is not definitive, the claim that devices have entered a topological regime weakens.
A second comment, arXiv preprint 2503.08944, engages directly with the Nature parity-readout paper. It offers alternative interpretations of the parity data, questioning whether the measurements confirm Majorana zero modes or could instead be explained by non-topological effects in the same device geometry. The distinction matters because a parity signal alone does not guarantee that the underlying physics is topological; conventional Andreev bound states, for example, can produce similar readout signatures under some conditions.
Microsoft responded in April 2025 with arXiv preprint 2504.13240, a formal reply addressing both comments. The response defends the methodology used in the Physical Review B and Nature papers while acknowledging specific points raised by the critics. The exchange is ongoing and has not yet been resolved through additional peer review or independent experiments. Until it is, the scientific community lacks consensus on whether the data establish Majorana zero modes or leave room for alternative explanations.
No public timeline or resource allocation figures exist for scaling Majorana 1 beyond its current form. Microsoft has not stated how many qubits the chip contains, what error rates it achieves in operation, or when a multi-qubit demonstration might follow. Competing approaches from companies using superconducting transmon qubits and trapped-ion systems already publish multi-qubit benchmarks, gate fidelities, and error-correction results, giving those road maps a transparency advantage that Microsoft’s topological program has yet to match.
How to read the evidence
The two peer-reviewed papers, one in Nature and one in Physical Review B, constitute primary evidence. Both underwent formal review, and the Nature paper provides open data and code for scrutiny. These are the strongest anchors for any claim about what Microsoft has demonstrated in the lab. The arXiv comments and Microsoft’s arXiv response are also primary documents, but they have not passed peer review, so their technical arguments carry less institutional weight. They do, however, reflect active scientific debate by researchers with domain expertise, and they identify specific methodological questions that peer review of the original papers may not have fully resolved.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
