Last updated: March 2026 | This article contains no affiliate links.

Microsoft Majorana 1: The Quantum Chip That Could Change Everything — Or Nothing

As of March 2026, quantum computing has three horses in the race: Google’s brute-force superconducting approach (105-qubit Willow), IBM’s industrial-scale strategy (1,121-qubit Condor), and Microsoft’s long-shot bet — a fundamentally different qubit architecture that, if it works, could make both competitors obsolete. That bet has a name: Majorana 1.

Microsoft announced the chip on February 19, 2025. The company calls it the world’s first Quantum Processing Unit powered by a Topological Core. The claim is bold. The science is contested. And in 2026, with DARPA now involved and a prototype deadline on the horizon, the pressure is real.

Background: Why Topological Qubits?

Every qubit — the fundamental unit of quantum computation — has a noise problem. Quantum states are fragile. Vibrations, electromagnetic radiation, temperature fluctuations: all of these destroy the coherence that makes quantum computing useful. IBM and Google solve this with software. They use error-correcting codes that require thousands of physical qubits to protect a single “logical” qubit you can actually compute with.

Microsoft is betting on a different approach: fix the hardware so you need far less error correction in the first place.

Topological qubits store quantum information in the shape of a quantum state rather than the state of a single particle. Think of it this way: you can crumple a piece of paper (local noise), but you cannot change whether it has a hole in it without tearing it (topological change). The idea is that local disturbances — the main killer of standard qubits — physically cannot flip a topological qubit’s state.

The specific particles Microsoft uses are called Majorana zero modes. They are exotic quantum objects that were first theorized by Italian physicist Ettore Majorana in 1937. For decades, physicists debated whether they existed at all. Microsoft has spent over 15 years and more than $1 billion trying to build a device that reliably creates and controls them.

Majorana 1 is the first chip they claim achieves this.

Technical Details: What Makes Majorana 1 Different

The chip is built on what Microsoft calls a “topoconductor” — a new class of material composed of an indium arsenide and aluminum heterostructure. The combination creates the conditions for Majorana zero modes to emerge at the edges of semiconductor nanowires cooled to near absolute zero.

Key specifications as of March 2026:

• Current qubit count: 8 topological qubits on the chip
• Error rate: Approximately 1% in initial measurements, with state-flip errors occurring only once per millisecond on average
• Qubit size: Roughly 10 microns per qubit — small enough to support the eventual million-qubit density target
• Control: Digital control electronics integrated directly on-chip, a significant engineering achievement at cryogenic temperatures
• Scale target: 1 million qubits on a single chip the size of a graham cracker (long-term roadmap)

Chetan Nayak, Microsoft’s Technical Fellow for Quantum Hardware, put the scale argument directly: “Whatever you’re doing in the quantum space needs to have a path to a million qubits. If it doesn’t, you’re going to hit a wall. We have actually worked out a path to a million.”

The architecture matters because of the math. Under IBM’s approach, you need roughly 1,000 physical superconducting qubits to produce one reliable logical qubit. Microsoft claims that topological qubits, due to their inherent error resistance, would require far fewer physical qubits per logical qubit. If true, the economics flip entirely.

Microsoft was also selected for the final phase of DARPA’s Underexplored Systems for Utility-Scale Quantum Computing (US2QC) program — a government benchmarking initiative that specifically targets approaches outside the mainstream. DARPA’s mandate to Microsoft: build a fault-tolerant prototype in years, not decades.

The Scientific Controversy No One Wants to Talk About

Here is what most tech coverage of Majorana 1 glosses over: the foundational Nature paper (Nature 638, 651–655, 2025) that Microsoft used to back their February 2025 announcement came with something genuinely unusual attached — an editorial disclaimer from the Nature team itself.

The disclaimer reads: “The editorial team wishes to point out that the results in this manuscript do not represent evidence for the presence of Majorana zero modes in the reported devices.”

This is not a standard caveat. Scientific journals rarely attach disclaimers to papers they publish.

The peer-review record makes the picture clearer. According to publicly available review reports, 3 of 4 initial reviewers rejected the paper. Two of those four maintained a “do not recommend” position through the final round, arguing the conclusions rested on “questionable hypothesis and methodologies.”

The core objection comes from condensed matter physicist Henry F. Legg, who submitted a formal challenge to arXiv on March 11, 2025 (arXiv:2503.08944). Analyzing Microsoft’s own public data repository, Legg found that the nanowire regions where Microsoft claimed to perform “parity readout” — the key measurement proving topological qubit behavior — were “highly disordered and present no clear gap.” His conclusion: “That these regions are gapless contradicts the claim that the reported measurements are of the parity of a superconducting nanowire… the core findings in Nature 638, 651-655 (2025) are not reliable and should be revisited.”

This is not the first time Microsoft has faced this problem. In 2018, the company published a Nature paper claiming evidence of Majorana particles. That paper was retracted in 2021 after independent researchers identified flaws in the data analysis. Professor George Booth of King’s College London noted that this history is precisely why the physics community is demanding more: Microsoft has “been burnt before,” and “there is still some healthy scepticism of the timescales for the roadmaps.”

Professor Paul Stevenson of the University of Surrey offered a calibrated assessment at the February 2025 announcement: “Their latest result shows that they have managed to build roughly one half of one qubit. Now the challenge is to build that up first into a single qubit, then an array of qubits. Until the next steps have been achieved, it is too soon to be anything more than cautiously optimistic.”

Microsoft’s transparency strategy — open-sourcing their data on public repositories — has produced an ironic side effect. The same data they published to build trust is now being analyzed by independent researchers to challenge their claims. It is accountability in both directions.

Industry Reactions: How Google, IBM, and Amazon Are Responding

The three major quantum competitors have each taken a different posture toward Microsoft’s topological bet.

Google is not pivoting. The company’s Willow chip (105 qubits, demonstrated in late 2024) uses superconducting transmon qubits. Google’s approach prioritizes demonstrating quantum advantage on specific benchmarks now, rather than waiting for a new qubit architecture to mature. Their logic: incremental progress with proven hardware beats theoretical superiority with unproven hardware.

IBM has doubled down on scale. The Condor chip at 1,121 qubits uses the same superconducting approach, and IBM has already transitioned manufacturing to 300mm semiconductor wafer fabrication — the same industrial-scale process used to make classical computer chips. IBM has publicly targeted verified quantum advantage by end of 2026. They are not betting on topological qubits.

Amazon announced its own Ocelot chip in February 2025, using a “cat qubit” approach that also aims at hardware-level error reduction. Amazon’s strategy is closer in philosophy to Microsoft’s, but uses a different physical mechanism. Neither company has announced timelines for commercial availability.

The key tension in the industry as of 2026: IBM and Google are racing on a known track, accepting today’s hardware limitations. Microsoft is building a different track and betting it will be faster once complete. The risk profile is entirely different.

What This Means for Developers, Enterprises, and Governments

For most developers and enterprises in 2026, the honest answer is: not much yet, practically speaking.

Microsoft has not announced a commercial availability date for Majorana 1. The chip is a research milestone, not a product. IBM’s quantum systems are accessible today via IBM Quantum Network, with cloud-based access to real superconducting hardware. Google offers quantum cloud access through Google Cloud. Microsoft’s Azure Quantum platform currently gives access to partner hardware — IonQ, Quantinuum — not Majorana 1.

The sectors paying close attention are defense, pharmaceuticals, and financial modeling. These are the industries where quantum computers that can run algorithms classical computers cannot — breaking RSA encryption, simulating molecular structures for drug discovery, optimizing portfolio risk at speeds no classical machine can match — would have immediate, measurable value.

Governments are ahead of enterprises on this. The US DARPA investment in Microsoft’s roadmap signals Washington’s concern about quantum advantage in national security contexts. The EU’s Quantum Flagship program is funding multiple competing architectures. China has announced its own topological qubit research programs, though verification of their progress is limited.

For enterprises building long-horizon tech strategy: the correct interpretation of Majorana 1 is not “quantum is here.” It is “Microsoft has a plausible path that, if validated, changes the economics of fault-tolerant quantum computing.” That path could be validated in 3 years or abandoned in 3 years. Both outcomes are possible.

What’s Next: DARPA Deadline and the 2029 Target

Microsoft’s published roadmap points to utility-scale quantum computing by 2029. “Utility-scale” in this context means a machine capable of solving problems that have no classical solution in any reasonable timeframe — not just faster than classical computers, but categorically beyond them.

The near-term milestones to watch in 2026 and 2027:

1. First full logical qubit: Microsoft needs to demonstrate a single, fully controllable logical qubit using topological hardware. This is the step Professor Stevenson flagged as the next required proof point.
2. Braiding demonstration: The quantum operation unique to topological qubits — “braiding” Majorana particles to perform logic gates — has not yet been demonstrated. This is the critical missing step between what Majorana 1 has achieved and a functioning topological qubit.
3. DARPA prototype delivery: Microsoft must deliver results to DARPA’s US2QC program. The specific benchmarks are not fully public, but the timeline is measured in years, not decades — meaning 2027–2028.
4. Independent verification: For any of Microsoft’s claims to carry broad scientific weight, they need independent lab replication. Given the 2021 retraction and the current arXiv challenge, replication will be closely watched.

Microsoft launched its Quantum Pioneers Program (QuaPP) in January 2026, focused on bridging theoretical physics with practical hardware through measurement-based quantum computing (MBQC). The program signals the company is building an ecosystem — researchers, engineers, partner institutions — around the Majorana architecture before it has fully proven itself.

The quantum computing race in 2026 is not a single race. It is three separate bets: IBM’s industrial-scale known quantity, Google’s benchmark-first approach, and Microsoft’s long-shot that could win everything or prove unworkable. The next 18 months of experimental data will determine which bet is paying off.

Frequently Asked Questions