The Quantum Investor's Notebook logo The Quantum Investor's Notebook
← Back

Sunday, June 28, 2026

The Majorana Qubit: A Bit Hidden Where Noise Cannot Reach

Microsoft’s bet is a qubitThe basic unit of a quantum computer. Like a 'bit' in a normal computer, but instead of being only 0 or 1 it can be 0, 1, or a blend of both at once. that stores its information in two places at once, so local noise has a much harder time reading the whole of it.

What a Majorana qubit is

Start with the problem every qubit has. Its information sits in one place, a spin, a current, a charge, and anything that can touch that place can corrupt it. The Majorana qubitA qubit that stores its information non-locally across two or more Majorana zero modes, so local noise has a much harder time reading or corrupting it. Microsoft's topological-qubit bet. does the opposite. It splits one bit across two points far apart and stores it in the relationship between them, not in either one alone.1

Hit one end all you like. The answer is not kept at either end.

Those two points are Majorana zero modesAn exotic, hard-to-create quantum state that Microsoft's topological qubit depends on. Its existence in these devices is still scientifically disputed., strange half-particle states that, in the right material, sit at the two ends of a wire. Alone, each is nothing you can read. In the simplest picture, the protected information lives in their shared parityWhether the number of particles in a system is even or odd. It is the quantity Microsoft's qubit reads out., whether the pair carries an even or an odd number of electrons.2 Noise striking one end cannot learn the full shared answer on its own, because the information is not stored at either end alone. To damage the protected information, ordinary noise has to disturb the shared state, not just poke one end. That non-locality is what topological protectionShielding quantum information by storing it non-locally, so local noise has a much harder time corrupting it. The further apart the parts sit, the safer it can get. means.

How you build and run one

Building this is synthetic physics; nature does not hand you Majorana modes. You take a very pure semiconductor nanowireAn extremely thin wire, billionths of a meter wide, used as the body of Microsoft's qubit device., lay it against a superconductor so it borrows superconductivity (the proximity effectWhen a superconductor placed against another material lends it superconducting behaviour it would not have on its own.), add a magnetic field, and tune the whole thing into the topological phase that the topological gap protocolMicrosoft's pre-registered pass/fail test for deciding whether a device has truly entered a topological state hosting Majorana modes, built to reject look-alike false signals. Its reliability is still debated. exists to certify.2

Operating it is the clever part. The textbook method, braidingMoving defects in a 2D code around one another so the path itself performs a protected logical operation., means physically swapping the modes around one another, which is brutally hard. Microsoft’s design instead groups four modes into one qubit and runs the logic by measurementReading a quantum system, which forces it out of its blend of possibilities into a single definite 0 or 1 and ends its quantum behaviour. alone, reading parities rather than moving anything.3 These are non-Abelian anyonsExotic particle-like objects whose record of being moved around one another changes the system's state, which is how a topological qubit would store and process information., whose very history of swaps changes the state.

A wire on a superconductor with a Majorana zero mode drawn as a star at each end. A dashed bracket spanning both ends labels the qubit information stored in their shared parity. A red noise arrow strikes one end and is stopped by a red X, showing that one end alone cannot read it.
Figure 1: Why it is protected. One bit lives in the shared parity of the two Majorana modes at the wire’s ends, so noise reaching only one end has a much harder time reading or flipping it.

Why it is so hard

The enemy is quasiparticle poisoningA stray electron sneaking into the device and flipping the qubit's parity, the main error a Majorana qubit has to fend off.: a single stray electron wandering in and flipping the parityWhether the number of particles in a system is even or odd. It is the quantity Microsoft's qubit reads out., erasing the bit.3 Worse, the defenses fight each other. Spreading the modes farther apart makes the qubitThe basic unit of a quantum computer. Like a 'bit' in a normal computer, but instead of being only 0 or 1 it can be 0, 1, or a blend of both at once. safer from noise but weakens the charging energy that keeps stray electrons out, so the design is a hunt for the balance.3

Hardest to build first, on a wager that it is easiest to scale later.

And the headline question, whether these modes truly exist in the devices, is still open. The signatures used to claim them can be mimicked by ordinary disorder, and in 2025 Microsoft reported reading a device’s parity in a single shot at about a 1 percent error.4 That result is now part of an active scientific dispute: a published Matters Arising article challenged the interpretation, and Microsoft published a reply defending it. I will be honest: this qubit is a bet, not a product.

The overhead bet

Here is why the bet is worth watching. Every other platform pays a brutal tax for reliability: an ordinary qubit is so fragile that error correctionTechniques that combine many shaky physical qubits into fewer reliable ones, so a long calculation stays correct. may need thousands of physical qubitsAn actual piece of qubit hardware. On its own it is fragile and makes frequent errors. to make one good logical one. A Majorana qubitA qubit that stores its information non-locally across two or more Majorana zero modes, so local noise has a much harder time reading or corrupting it. Microsoft's topological-qubit bet. would arrive partly protected by its own physics, needing far less of that overhead.2

So the trade is stark. Microsoft is choosing the qubit hardest to build first, betting it is easiest to scale later. If the materials cooperate and the modes are real, the path to a million-qubit machine gets shorter. If they are not, it is a decade spent chasing a particle that was never there. Both endings are still open.

Sources

  1. Kitaev, A. Yu. “Unpaired Majorana Fermions in Quantum Wires.” Physics-Uspekhi 44, 131–136 (2001). DOI: 10.1070/1063-7869/44/10S/S29. Preprint: arXiv:cond-mat/0010440.
  2. Das Sarma, S., Freedman, M. & Nayak, C. “Majorana Zero Modes and Topological Quantum Computation.” npj Quantum Information 1, 15001 (2015). DOI: 10.1038/npjqi.2015.1. Preprint: arXiv:1501.02813.
  3. Karzig, T., Knapp, C., Lutchyn, R. M. et al. “Scalable Designs for Quasiparticle-Poisoning-Protected Topological Quantum Computation with Majorana Zero Modes.” Physical Review B 95, 235305 (2017). DOI: 10.1103/PhysRevB.95.235305. Preprint: arXiv:1610.05289.
  4. Microsoft Quantum. “Interferometric Single-Shot Parity Measurement in InAs-Al Hybrid Devices.” Nature 638, 651–655 (2025). DOI: 10.1038/s41586-024-08445-2. The interpretation was challenged in a published Matters Arising (Legg, H. F., “On the Robustness of Topological Gap Detection via Transport,” Nature 654, E22–E26, 2026; preprint arXiv:2503.08944), with a Reply from Microsoft Quantum in the same issue.