TL;DR
- Majorana 2: Microsoft has unveiled Majorana 2, the successor to its Majorana 1 quantum chip, as a topological processor built for longer-lived qubits.
- Reliability Figure: Microsoft says the new chip raises qubit reliability 1,000-fold and supports its 2029 quantum-computer roadmap.
- Evidence Gap: Outside physicists say the preprint and limited device data do not yet show repeatable results at scale.
- Proof Tests: Peer review, multi-device reproduction, and DARPA evaluation will test whether the hardware can back the roadmap.
Microsoft has unveiled Majorana 2, a topological quantum processor that follows its 2025 Majorana 1 chip and is meant to push qubit stability closer to a scalable machine. The device gives Microsoft a new hardware milestone, but the larger test is whether its reliability figures can move from promising lab result to repeatable quantum-computing evidence.
Microsoft’s current 1,000-fold reliability figure now anchors the chip’s significance. Its hardware roadmap points to a scalable quantum computer anticipated by 2029, a target that now sits beside IBM’s 2025 rival 2029 quantum roadmap. Repeatable reliability evidence sits at the center of the roadmap rather than a narrow lab metric.
Microsoft’s topological approach is meant to make qubits less vulnerable to local noise. The processor is less a settled breakthrough than a reproducibility test for Microsoft’s quantum roadmap.
What Microsoft Says Changed Inside the Chip
Majorana 2 changes the physical stack behind Microsoft’s topological-qubit approach with a lead-based superconducting layer instead of aluminum. Its active semiconductor region uses indium arsenide with indium arsenide antimonide. In plain terms, Microsoft is trying to build qubits that are less vulnerable to the interference that can quickly destroy quantum information.
A wider protective energy separation known as the topological gap is central to the proposed improvement. In Microsoft’s figures, the new stack has a more-than-double topological gap compared with its previous processor.
Majorana 2 qubit lifetimes exceed 20 seconds and sometimes exceed one minute in Microsoft’s figures. Majorana 1, by comparison, sat in the one-to-12-millisecond range. Independent testing still has to show whether that before-and-after jump holds beyond Microsoft’s own devices.
Microsoft’s 2025 Majorana 1 predecessor framed topological qubits as a stability path, while older Majorana fermion work explains why Microsoft has kept pursuing that design since 2018. Majorana 2 raises the stakes by attaching much larger reliability numbers to the same route.
Why the Evidence Is Still Contested
Majorana quasiparticles could be more resistant to physical noise than ordinary qubits. That theory does not remove the burden to show repeated, device-level results across enough hardware to prove the effect is not a one-off artifact.
Henry Legg, a physicist at the University of St. Andrews, says the preprint’s limited device evidence leaves Microsoft’s reproducibility claim unresolved:
“You can see something amazing in one device and never see it again because it’s just some random artifact.”
Henry Legg, physicist at the University of St. Andrews (via Scientific American)
Legg also questions whether the work would meet ordinary peer-review expectations. Microsoft’s current claim rests on a new preprint manuscript that had not been peer reviewed on publication day, so the reliability figures still need journal review and independent replication before they can be treated as settled performance evidence.
Sergey Frolov, a quantum researcher at the University of Pittsburgh, went further, arguing that Microsoft’s latest preprint depends on a research record he does not consider firm enough:
“This new preprint is not based on a research track record that can be considered a solid foundation.”
Sergey Frolov, quantum researcher at the University of Pittsburgh (via Scientific American)
Frolov’s critique matters because Microsoft’s 2029 target depends on the same claim becoming repeatable hardware evidence. A strong preprint can guide the field, but peer review and multi-device reproduction will decide whether Majorana 2 moves beyond a promising company result.
Rivals and the Next Proof Point
Microsoft is not the only company trying to make quantum hardware reliable enough for practical use. Rivals are purs
uing different designs, including neutral-atom systems, superconducting chips, annealing machines, and photonic approaches. Amazon’s Ocelot chip and IBM’s 2029 quantum roadmap put Microsoft’s target inside a broader race to turn lab progress into deployable quantum systems.
Competing approaches do not settle Microsoft’s claim, but each vendor must show better qubits and a credible path to scaling them. Microsoft’s topological route promises durability if the physics holds up, while other vendors are pursuing neutral-atom, superconducting, annealing, and photonic systems that expose different engineering trade-offs.
Microsoft says its quantum team used AI to manage workflows, automate measurements, optimize fabrication, and find flaws. Those tasks matter because topological hardware depends on repeatable materials work and measurement cycles, not just a single impressive device result.
The US Defense Advanced Research Projects Agency (DARPA) has advanced Microsoft to the final phase of a utility-scale quantum computing evaluation program, giving the roadmap an external review path without validating Majorana 2’s performance claims by itself. The preprint still needs peer review, and multi-device reproducibility will determine whether Majorana 2 can support Microsoft’s 2029 roadmap as hardware evidence rather than a company target.
