Quantum Computing Modalities: Topological Quantum Computing
Table of Contents
(For other quantum computing modalities and architectures, see Taxonomy of Quantum Computing: Modalities & Architectures)
What It Is
Topological quantum computing encodes quantum information in exotic quasiparticles called non-Abelian anyons, whose quantum states are protected by the global topology of the system rather than by the local properties of any individual component. The core idea: if quantum information is stored in the way quasiparticles are arranged relative to each other (their topological configuration), then local noise, which affects individual particles but cannot change the global arrangement, cannot corrupt the computation. Gates are performed by braiding these quasiparticles around each other, and the resulting quantum operation depends only on the topology of the braid (which paths crossed over which), not on the speed, precision, or path details of the physical motion.
This is the most theoretically compelling approach to quantum computing. If topological qubits work as the theory predicts, they would be inherently fault-tolerant at the hardware level, reducing or eliminating the massive error-correction overhead that every other modality must carry. Where a superconducting surface-code architecture needs ~1,000 physical qubits per logical qubit, a topological qubit could in principle be its own logical qubit, or close to it. Microsoft has stated that its topological approach targets a hardware error rate of ~10⁻⁶ per operation, compared to the ~10⁻³ typical of today’s best physical qubits in other modalities.
The problem is that no one has demonstrated a topological qubit. The theory is beautiful. The physics has not cooperated.
How It Works
Non-Abelian Anyons and Braiding
In two-dimensional quantum systems, particles can exhibit exchange statistics that are neither bosonic nor fermionic. These “anyons” acquire a phase (Abelian anyons) or undergo a unitary transformation (non-Abelian anyons) when exchanged. The non-Abelian case is what matters for quantum computing: exchanging (braiding) two non-Abelian anyons applies a quantum gate to the system’s state that depends only on the topological class of the braid, not on the details of the exchange path.
The leading candidate for non-Abelian anyons in a solid-state system is the Majorana zero mode (MZM): a quasiparticle that appears at the boundary of a topological superconductor. Majorana modes come in pairs. A pair of spatially separated Majorana modes encodes a single fermionic mode, which can be occupied (|1⟩) or unoccupied (|0⟩). This occupation is the qubit. Because the two Majorana modes are spatially separated (potentially by micrometers), local noise at either site cannot determine or flip the occupation state. Only a process that connects both sites (like braiding one Majorana around the other) can change the qubit.
Braiding Majorana modes implements Clifford gates (rotations that, combined, can perform many useful operations but cannot alone achieve universal quantum computation). For universal computation, a non-Clifford gate (like a T-gate or π/8 rotation) is also needed. This requires either magic state distillation (reintroducing some error-correction overhead, though far less than in non-topological systems) or a different class of anyons (Fibonacci anyons, covered in my Fibonacci anyons deep dive, which are computationally universal by braiding alone but even more experimentally challenging to create).
The Topological Gap
The energy difference between the topological ground states (where the qubit information lives) and excited states (where errors can occur) is called the topological gap. For topological protection to work, this gap must be large compared to the thermal energy (kT) and any environmental perturbation energy. At millikelvin temperatures (~10 mK), the thermal energy is ~1 µeV, so the topological gap needs to be at least 10–100 µeV to provide meaningful protection. Measuring and maximizing this gap has been a central experimental challenge.
Material Platforms
The primary material system for Majorana-based topological qubits is a semiconductor-superconductor heterostructure: a semiconductor nanowire (typically InAs or InSb, chosen for their strong spin-orbit coupling) in proximity to a superconductor (aluminum). Under an applied magnetic field, theory predicts that the nanowire enters a topological superconducting phase, with Majorana zero modes appearing at the wire’s ends.
Microsoft‘s current approach uses InAs/Al heterostructures in a “topoconductor” geometry, where the semiconductor-superconductor interface is engineered for cleaner proximity-induced superconductivity than previous nanowire designs.
Alternative platforms that could host non-Abelian anyons include fractional quantum Hall systems (the ν = 5/2 state is theorized to host Ising anyons), vortex cores in topological superconductors, and engineered lattice models. None of these has demonstrated a usable topological qubit.
Key Academic Papers
Kitaev (2003). “Fault-tolerant quantum computation by anyons.” The foundational theory paper proposing that non-Abelian anyons could implement fault-tolerant quantum gates through braiding. Introduced the toric code as a concrete model. Published in Annals of Physics.
Nayak, Simon, Stern, Freedman, Das Sarma (2008). “Non-Abelian anyons and topological quantum computation.” The comprehensive review that defined the field, covering anyon theory, candidate physical systems, and computational capabilities. Published in Reviews of Modern Physics. The standard reference for anyone entering the field.
Mourik et al. / Kouwenhoven (2012). Observed a zero-bias conductance peak in InSb nanowires proximitized with NbTiN superconductor, consistent with Majorana zero modes. Published in Science. This generated enormous excitement but also decades of controversy: subsequent work showed that zero-bias peaks can arise from non-topological disorder effects (Andreev bound states, smooth confinement potentials), making the signature ambiguous.
Retraction of the 2018 quantized conductance claim. In 2021, a high-profile Nature paper from Microsoft-affiliated researchers (Delft) claiming quantized Majorana conductance was retracted after independent scrutiny revealed data analysis concerns. This episode significantly damaged confidence in the experimental Majorana program and raised the evidentiary bar for subsequent claims. (The original retracted paper is at https://doi.org/10.1038/nature26142.)
Microsoft Majorana 1 (February 2025). Microsoft announced an 8-qubit-capacity test chip in an InAs/Al topoconductor platform, with a companion paper published in Nature. My analysis. The peer-reviewed evidence does not demonstrate a topological qubit. Nature reviewers explicitly noted that the published measurements “do not, by themselves, determine whether the low-energy states detected by interferometry are topological.” Additional data presented at APS March Meeting 2025 met substantial skepticism from the condensed-matter physics community (Sergey Frolov, Patrick Lee, Henry Legg raised specific objections). Detailed data analysis.
Australian 1/f noise preprint (June 2025). A preprint from Australian researchers raised a fundamental concern about 1/f noise decoherence of Majorana qubits, suggesting that charge noise in the semiconductor-superconductor system could decohere the topological states faster than previously assumed. If this analysis holds, it would undermine the thesis that topological qubits are inherently more stable than conventional qubits.
Where Things Actually Stand (May 2026)
This section is where I depart from the theoretical promise and assess the evidence as it exists today.
No topological qubit has been demonstrated. Despite over two decades of theoretical development and more than a decade of experimental effort, no group has demonstrated a qubit whose information is stored in a topological degree of freedom. Microsoft’s Majorana 1 is an 8-qubit-capacity test chip, but “capacity” means the chip has the physical geometry to host 8 Majorana pairs. It does not mean 8 topological qubits have been created, controlled, or measured. The Nature reviewers’ statement is the clearest summary: the measurements do not determine whether the observed states are topological.
No two-qubit gate has been demonstrated. The most basic requirement for a quantum computer is a controllable two-qubit interaction. No braiding operation between Majorana modes has been experimentally demonstrated. The gap between “observed a zero-bias peak consistent with a Majorana” and “braided two Majoranas to implement a CZ gate with measured fidelity” is enormous.
The evidentiary record is contested. The 2021 retraction of the Delft quantized conductance paper, combined with ongoing debate about whether observed zero-bias peaks are truly topological (as opposed to trivial Andreev states), means that even the existence of Majorana zero modes in these devices remains under active scientific dispute. This is not a case where the physics is settled and only the engineering remains. The physics itself is unresolved.
Microsoft is hedging. Microsoft’s partnership with Atom Computing to deliver the Magne system (50 logical qubits from ~1,200 neutral-atom physical qubits) to Denmark by early 2027 is a pragmatic acknowledgment that topological qubits will not deliver commercial quantum computing on Microsoft’s Azure timeline. Microsoft continues to fund the topological program through DARPA US2QC and its Azure Quantum roadmap, but it is building its near-term quantum business on neutral atoms.
Comparison to Other Modalities
The comparison is brief because topological qubits exist in a different category from the other modalities in this series. Superconducting, trapped-ion, neutral-atom, silicon-spin, and photonic systems have all demonstrated working qubits, multi-qubit gates, and (for the first three) error-corrected logical qubits. Topological systems have demonstrated none of these.
The value proposition, if realized, would be transformative: topological qubits with ~10⁻⁶ hardware error rates would need minimal error-correction overhead, potentially encoding a logical qubit in a single physical qubit or a small number of anyons. Compare this to the ~1,000:1 overhead for superconducting surface codes, the ~2:1 for trapped-ion or neutral-atom qLDPC codes, and the even higher overhead for photonic FBQC architectures. If Microsoft’s vision is correct, a few thousand topological qubits could match what other modalities need millions of physical qubits to achieve.
But “if realized” is doing extraordinary work in that sentence. Every other modality in this series has crossed the threshold from theory to experimental demonstration. Topological has not. The comparison is between a proven but inefficient approach (active error correction on imperfect physical qubits) and an unproven but potentially efficient approach (hardware-level topological protection). As of May 2026, the proven approach is winning, and it is winning by such a large margin that topological qubits are not a factor in any near-term CRQC timeline assessment.
Advantages (Theoretical)
I label these “theoretical” because none has been experimentally demonstrated.
Hardware-level error protection. If topological qubits work as predicted, local noise cannot corrupt the quantum information because it is stored non-locally across spatially separated anyons. This would reduce error rates by orders of magnitude compared to conventional physical qubits, potentially eliminating most error-correction overhead.
Topological gate stability. Braiding operations produce quantum gates whose outcome depends only on the topology of the braid path, not on speed, timing, or path details. Small imperfections in the physical motion of anyons do not introduce gate errors. This is analogous to how a knot’s identity does not depend on the exact shape of the rope.
Reduced overhead for fault tolerance. If per-gate error rates are ~10⁻⁶ instead of ~10⁻³, the number of physical qubits needed per logical qubit drops from ~1,000 (surface code) to potentially 1–10. This could make fault-tolerant quantum computing practical with far fewer total qubits.
Disadvantages (Demonstrated)
No demonstrated qubit. The most fundamental disadvantage. Over 20 years of effort have not produced a single controllable topological qubit.
Contested evidence for Majorana zero modes. The experimental signatures claimed to indicate Majorana modes (zero-bias conductance peaks) can be produced by non-topological mechanisms. The 2021 retraction damaged the field’s credibility. The Australian 1/f noise preprint (July 2025) raised additional concerns about decoherence.
No two-qubit gate, no braiding demonstration. The entire computational paradigm (braiding anyons to implement gates) has not been experimentally demonstrated.
Limited universality from Majorana braiding. Even if Majorana braiding works, it only produces Clifford gates. Universal computation requires non-Clifford gates (T-gates), which must be implemented through magic state distillation or non-topological operations. This reintroduces some of the error-correction overhead that the topological approach aims to eliminate.
Extreme material requirements. Topological superconductivity requires ultra-clean semiconductor-superconductor interfaces, specific magnetic fields, and millikelvin temperatures. The material science challenges are at least as demanding as any other modality, and arguably more so.
Single vendor. Microsoft is effectively the only major organization pursuing Majorana-based topological qubits for quantum computing. This creates concentration risk: if Microsoft deprioritizes the program (as some read the Atom Computing partnership), the entire topological qubits research pipeline narrows significantly.
Impact on Cybersecurity
Topological qubits do not currently affect CRQC timeline assessments. They are not included in any credible Q-Day prediction model because there is no demonstrated topological qubit from which to extrapolate a scaling trajectory.
If topological qubits were to work as theorized, the implications would be significant. A topological quantum computer with ~10⁻⁶ error rates per gate could run Shor’s algorithm on RSA-2048 with far fewer physical qubits than any other approach, potentially with only a few thousand topological qubits rather than millions of conventional physical qubits. This would make CRQCs smaller, cheaper, and more accessible. Microsoft has stated a goal of reaching fault-tolerant quantum computing within this decade via the topological approach.
But this is a conditional statement about a technology that has not demonstrated its most basic building block. I do not include topological qubits in my CRQC Quantum Capability Framework assessment of current progress toward a CRQC. If a peer-reviewed demonstration of a topological qubit with measured gate fidelity is published, I will reassess.
The practical guidance remains: PQC migration timelines should not be relaxed based on topological qubits’ theoretical promise. They should not be accelerated based on it either. The technology is too uncertain to influence planning in either direction. Regulators and clients are setting deadlines based on demonstrated progress in superconducting, trapped-ion, and neutral-atom systems, which is the right basis for decision-making.
Future Outlook
What would change the assessment. A peer-reviewed demonstration of a two-qubit topological gate with measured fidelity would be the single most important milestone. Until that happens, topological quantum computing remains a research program, not a technology on a deployment trajectory. More specifically, what I would need to see:
- A controlled braiding operation between two Majorana zero modes, with the resulting quantum state measured and confirmed to match the predicted topological transformation.
- A measured gate fidelity that demonstrates the theorized noise suppression (i.e., fidelity significantly higher than what the same material system achieves with non-topological qubits).
- Replication by an independent group.
Microsoft’s stated timeline. Microsoft’s goal is fault-tolerant quantum computing within this decade via the topological approach. The company remains in DARPA’s US2QC portfolio. Whether this timeline is realistic given the current state of the evidence is, to put it diplomatically, uncertain. Microsoft’s Azure Quantum business is being built on Atom Computing neutral atoms in the near term, with topological qubits as the long-term aspiration.
The broader topological research ecosystem. Academic groups at Delft/QuTech, Copenhagen (Niels Bohr Institute), Princeton, Caltech, and in Israel continue fundamental research on topological phases of matter, Majorana modes, and fractional quantum Hall anyons. IBM has reported interferometry results in fractional quantum Hall systems. These are fundamental physics programs, not engineering programs aimed at building computers.
My position. Topological qubits represent the highest-risk, highest-reward approach in quantum computing. If they work, they could leapfrog every other modality by eliminating the error-correction overhead that dominates all current scaling plans. The theoretical physics is compelling. The experimental evidence is not. Until it is, topological quantum computing belongs in the “track as strategic intelligence” category, not in procurement plans or CRQC timeline models.
Topological QC Sub-Modalities
This article covers topological quantum computing as an umbrella. Two sub-modalities have their own dedicated deep dives:
Majorana Qubits
The leading experimental approach to topological qubits. Majorana zero modes in semiconductor-superconductor heterostructures (InAs/Al, InSb/Al) are the quasiparticles that Microsoft and academic groups are attempting to create, control, and braid. The Majorana article covers the physics, the experimental history (including the contested evidence), and Microsoft’s specific hardware program in more detail than this umbrella article.
Fibonacci Anyons
A theoretically more powerful class of non-Abelian anyons whose braiding statistics alone can implement universal quantum computation (unlike Majorana/Ising anyons, which only produce Clifford gates). Fibonacci anyons are predicted to exist in certain fractional quantum Hall states (notably ν = 12/5) but are even further from experimental realization than Majorana modes. The Fibonacci article covers the mathematical framework, the computational universality proof, and the candidate physical systems.
Quantum Upside & Quantum Risk - Handled
My company - Applied Quantum - helps governments, enterprises, and investors prepare for both the upside and the risk of quantum technologies. We deliver concise board and investor briefings; demystify quantum computing, sensing, and communications; craft national and corporate strategies to capture advantage; and turn plans into delivery. We help you mitigate the quantum risk by executing crypto‑inventory, crypto‑agility implementation, PQC migration, and broader defenses against the quantum threat. We run vendor due diligence, proof‑of‑value pilots, standards and policy alignment, workforce training, and procurement support, then oversee implementation across your organization. Contact me if you want help.
