Microsoft

Overview

Microsoft takes a different path from most quantum companies. Rather than building the superconducting or trapped-ion qubits that other firms use, it has spent close to two decades trying to create a topological qubit, a type of qubit that would be naturally protected from many kinds of errors. If it works, it could be far easier to scale than today's noisy qubits. If it does not, Microsoft has hedged with a cloud platform and partnerships.

That hedge is Azure Quantum, a cloud service that gives developers access to quantum hardware and simulators from several providers, along with a growing collaboration with hardware companies such as Atom Computing and Quantinuum on error-corrected logical qubits.

History and background

Microsoft's quantum program grew out of Station Q, a research group set up in Santa Barbara, California in the mid-2000s and led by the mathematician Michael Freedman, a Fields Medal winner known for his work in topology. The idea was to turn an exotic theoretical concept, particles whose braiding patterns could store information, into working hardware. Microsoft also built the software side, releasing the Q# programming language and the Quantum Development Kit in 2017 and announcing the Azure Quantum cloud in 2019.

The hardware effort has been controversial. A 2018 Nature paper from a Microsoft-linked team in Delft claiming evidence for Majorana particles was retracted in 2021 after problems were found in the data analysis. Microsoft continued the program with new materials and measurement methods, and in February 2025 it announced Majorana 1, which it called the first quantum processor with a topological core. Independent physicists have said the evidence for genuine topological qubits is still incomplete.

How the technology works

A conventional qubit stores information in a fragile physical property, such as the state of a circuit or an ion. A topological qubit would store it in a collective property of the system that local disturbances cannot easily change. Microsoft's design uses semiconductor-superconductor nanowires made from a material stack it calls a topoconductor. At near absolute zero and in a magnetic field, the wire is supposed to host Majorana zero modes at its ends, and the qubit's state is encoded in the parity, even or odd, of the electrons shared between them.

Reading and controlling these qubits uses voltage pulses and measurements rather than the microwave pulses used with superconducting chips, which Microsoft argues makes the system digitally controlled and easier to scale. Microsoft has said Majorana 1 holds eight qubits on a design intended to accommodate up to one million.

The central scientific dispute is proof. When Majorana 1 was announced, the editorial notes accompanying the related Nature paper said the results did not by themselves demonstrate Majorana modes, and critics said Microsoft had not shown a working topological qubit. In June 2026 Microsoft announced Majorana 2 with a new material stack, reporting a parity lifetime of about 20 seconds and a topological gap more than twice as large as before. Nature's news coverage noted that some researchers remained sceptical.

Products and platforms

Key milestones

Strengths and challenges

Strengths

• A potentially very scalable qubit design, if topological protection works as hoped.
• A mature software and cloud business that does not depend on a single hardware bet.
• Long-term commitment: Microsoft has funded this research for close to two decades.
• Partnerships that deliver logical-qubit results now, on other companies' hardware.

Challenges and open questions

• Many independent researchers remain sceptical that Microsoft has demonstrated a fully working topological qubit.
• Majorana devices are far behind superconducting and trapped-ion systems in demonstrated computations.
• The 2021 retraction and the disputed 2025 claims mean Microsoft's announcements attract heavy scrutiny.
• The 2029 target for a scalable topological system is ambitious and depends on further materials progress.

Why it matters

Microsoft's bet is the highest-risk, highest-reward approach among the big technology companies. If topological qubits work, the number of physical qubits needed for a useful machine could drop dramatically, changing the economics of the entire field. That is why the debate over Majorana 1 and Majorana 2 matters even to people who never plan to use one.

For developers, Azure Quantum and its resource estimator are practical tools today. They help answer a question that matters for planning: how big must a quantum computer be before a given problem becomes solvable? Estimates like these are what make timelines for cryptography, chemistry, and finance more concrete.

Latest news

• Microsoft announces Majorana 2 at Build, reporting a 20-second parity lifetime and targeting a scalable topological system by 2029 — June 2026
• Nature reports that researchers remain sceptical of Microsoft's Majorana 2 claims — June 2026
• QuNorth says Magne, the 50-logical-qubit machine built with Atom Computing, is expected to be fully operational in early 2027 — 2026
• Microsoft expands its quantum fabrication and research lab in Denmark — November 2025

Frequently asked questions

This profile is written for educational purposes and summarizes public announcements, company filings, and published research. Last reviewed: September 30, 2026. It is not investment advice. Check the company's official channels for the latest figures and claims.

← Back to all quantum companies