Qubit Modalities: "Other" (Silicon, Diamond, Topological, etc.)

This is the fifth and final post in my series on qubit modalities. I’ve already covered Neutral Atoms, Trapped Ions, Superconducting and Photonics. If you’ve been following along, you know that each modality has its own physics, its own strengths, and its own set of hard engineering problems to solve. While those four modalities capture much of the capital and headlines, there are a number of other approaches that need to be in the conversation. These “Other” approaches are made up of silicon spin qubits, quantum dots, donor atoms in silicon, topological qubits, diamond vacancy systems, carbon nanotubes, and a handful of more specialized solid-state approaches. This is where you find Intel, Microsoft, Silicon Quantum Computing, Diraq, Quantum Motion, Quantum Brilliance, Quobly, Equal1, SemiQon, C12, and others.

While this is a diverse and disparate set of other approaches, one easy way to frame this category besides “Other” is to consider what the approach might look like if it is designed to look less like a physics experiment and more like a chip industry product. That is the center of gravity for many of these “other” modalities. Superconducting systems have made enormous progress, but they still require large cryogenic systems. Neutral atoms have beautiful scaling characteristics, but they are built around elaborate lasers, vacuum systems, atom control, and optical engineering. Trapped ions have exceptional qubit quality, but scaling and speed remain central questions. Photonics has a compelling networking and manufacturing story, but still has to wrestle with loss, sources, detectors, and resource overhead.

These “other” modalities often attack the problem from a different angle by asking the question: Can we make quantum hardware using the chip industry infrastructure, materials, and manufacturing discipline that already gave us the classical computing industry?

Silicon spin qubits and quantum dots want to ride semiconductor manufacturing. Donor atom approaches want to place atoms with almost surgical precision. Topological qubits want to make errors intrinsically harder by encoding information in exotic states of matter. Diamond vacancy systems want stable, deployable quantum processors that do not require refrigerator-sized cryogenic systems. Carbon nanotube approaches are betting that ultra-clean materials can give qubits a quieter environment.

In this post I’ll cover how these modalities work (at a high, mostly non-technical level), why investors and technical teams are excited, who the important companies, and some assessment of technical claims.

In order to level set these “Other” modalities with the prior posts in this series, the following table provides a very high-level description of each, along with some of the strengths and weaknesses:

1. Silicon Spin Qubits and Quantum Dots: “Quantum Transistors”

The most important subcategory here is silicon spin.

A silicon spin qubit uses the spin of an electron, generally whether it is pointing “up” or “down” in a magnetic sense, as the quantum information carrier. In many architectures, the electron is trapped inside a tiny engineered pocket called a quantum dot. Think of a quantum dot as a nanoscale parking spot for an electron. You build a device that can park one electron, control it, move it near another electron, and measure what happened.

The reason this is the top subcategory of this bunch is because silicon is not some exotic lab material, it’s the foundation of the semiconductor industry. If quantum dots can be made with semiconductor tools, the scaling story becomes much more familiar. We can start talking about 300-millimeter wafers, process control, yield, uniformity, cryogenic probing, and integration with control electronics which is all a language the chip industry already speaks.

Intel’s Tunnel Falls program, for example, used a 12-qubit silicon spin chip made available to university and federal labs, and Intel has also reported 300-millimeter wafer-scale spin-qubit work focused on uniformity, fidelity, and measurement statistics. Diraq and imec reported a particularly important result in 2025: silicon spin-qubit unit cells fabricated in a 300-millimeter foundry environment with single- and two-qubit control fidelities above 99% across four devices, and state-preparation-and-measurement fidelities up to 99.9%.

The takeaway is that foundry-made devices are starting to show performance that belongs in the same conversation as more artisanal academic devices. That is a meaningful shift.

2. Donor Atoms in Silicon: Atomic-scale manufacturing

Silicon Quantum Computing, the Australian company associated with Michelle Simmons and UNSW, takes a different silicon route. Instead of defining a quantum dot mostly with electrical gates, the company’s 14|15 platform uses phosphorus atoms in silicon. The idea is to place atoms with extraordinary precision, then use their nuclear and electron spins as qubits.

That atomic precision is both the attraction and the challenge. If you can place atoms exactly where you want them, you can get beautiful control and connectivity. But atomic-scale placement is also not the same thing as mainstream semiconductor throughput. It is more exacting, more specialized, and harder to imagine scaling quickly unless the manufacturing flow matures.

The technical evidence is real. In late 2025, a Nature paper reported an 11-qubit atom processor in silicon, with single- and multi-qubit gate fidelities ranging from 99.10% to 99.99%, Bell-state fidelities up to 99.5%, and entanglement maintained across up to eight nuclear spins. That is impressive. The open question is whether this can become an industrial platform rather than a precision scientific instrument.

3. Topological Qubits: Tying quantum information into a knot

Topological quantum computing is the most conceptually elegant and most controversial category in this post. The metaphor people often use is a knot or braid. In normal qubits, information is fragile: local noise can disturb it. In a topological qubit, the dream is to store information in a more global property of the system, so the environment has a harder time corrupting it. It is like the difference between smudging a pencil mark and untying a knot. Smudging is easy. Untying a knot requires more structured interference.

Microsoft has spent years pursuing this route using Majorana zero modes in engineered semiconductor-superconductor devices. In 2026, Microsoft announced Majorana 2, claiming a lead-based material stack, microsecond-scale operations, mean lifetimes of 20 seconds, occasional lifetimes beyond one minute, and a revised target of scalable quantum computing by 2029. If those claims are ultimately validated, they would be a very big deal. Topological qubits could reduce overhead and make error correction less punishing.

My view: topological qubits are not something to dismiss. But they are not something to underwrite on press releases alone either. I look forward to peer-reviewed data to further legitimize this modality.

4. Diamond Vacancies: Quantum computing without the refrigerator drama

Diamond vacancy systems are different again. A diamond vacancy is a defect in a diamond lattice — for example, a nitrogen-vacancy center, where a nitrogen atom sits next to a missing carbon atom. That defect can behave like a controllable quantum system. The nice feature is that diamond is an incredibly stable host material. Some diamond-based systems can operate at room temperature, avoiding one of the biggest infrastructure headaches in quantum hardware.

Quantum Brilliance is the best-known company here. Its pitch is not, “We have the highest qubit count in the world.” It is closer to: “We can build compact, deployable quantum accelerators that fit into existing computing environments.” Pawsey Supercomputing Research Centre installed a room-temperature, diamond-based Quantum Brilliance system on site in 2022, pairing it with classical HPC infrastructure. Oak Ridge later described a Quantum Brilliance system in its Advanced Computing Ecosystem testbed as a hybrid full-stack platform with QPU, GPU, and CPU components. However, the Oak Ridge set-up only had two-qubit QPUs which are not going to run useful industrial algorithms.

5. Carbon Nanotubes and Other Materials-first Bets

Then there are the longer-tail approaches: carbon nanotubes, electrons on helium, molecular spins, defects in other crystals, and various hybrid architectures.

C12 in France is building carbon-nanotube-based quantum processors, betting that an exceptionally clean material environment can reduce noise. The company raised an €18 million financing round in 2024 and describes itself as a spin-out of the École Normale Supérieure physics lab in Paris.

These approaches are earlier and less de-risked, but they are worth watching because quantum computing is still young enough that materials breakthroughs can change the trajectory and competitive landscape rather quickly.

Summary: If you remember only three things about these Other Modalities, remember these:

• Silicon is a manufacturing thesis as much as a physics thesis. The question is not just whether the qubits work; it is whether they can be made, measured, tuned, and controlled like an industrial product.

• Topological qubits are the highest-upside wildcard. If Microsoft is right, the architecture could leapfrog parts of the field. But the proof burden is unusually high.

• Diamond and other solid-state approaches may win first on deployment constraints, not headline qubit counts. A smaller quantum system that can live inside real HPC or edge environments can teach the market things that lab-bound systems cannot.\

The sector is still early, but it no longer feels purely academic. The more interesting companies are not just saying, “Here is a qubit.” They are saying, “Here is how this becomes a system.”

If I had to pick one company from this list that looks most likely to catch up to the primary modalities, I would pick Diraq — not because it has the loudest commercial story, but because its technical milestones line up with the right industry thesis.

Silicon spin has the most natural bridge into semiconductor manufacturing. Diraq has credible academic and industrial DNA. Its 2025 Nature result with imec showed high-fidelity spin-qubit unit cells made in a 300-millimeter foundry environment. And the company is now also tied into U.S. semiconductor policy through its announced CHIPS-related letter of intent. That does not mean Diraq is “ahead” of superconducting, neutral atoms, trapped ions, or photonics today. It is not. Those modalities have more visible cloud systems, larger public user ecosystems, and more demonstrated multi-qubit experiments at meaningful scale.

Milestones that would meaningfully de-risk or re-rate Diraq

• Larger two-dimensional array — not just a few excellent devices, but tens to hundreds of qubits with public data on yield, fidelities, readout, crosstalk, and calibration time.

• Parallel operation. Many qubits working one at a time is not the same as a processor. Diraq needs to show that high-quality operations survive when multiple gates and measurements happen across the chip.

• Error-correction primitive. Not a full fault-tolerant machine, but something like a small surface-code patch, repeated stabilizer measurement, or a logical memory experiment where the logical behavior improves in a way that outside experts find credible.

• Automation. The 2025 Nature paper itself notes that calibration at the demonstrated level remained intensive and manual, and that automation needs to mature. That is exactly the next frontier.

• Control integration. Silicon spin qubits are tiny, but the control problem is not tiny. A serious de-risking event would be evidence that cryogenic control, wiring, heat load, and signal integrity can scale without destroying qubit performance.

If I had to pick one player in this “other modalities” group that could most dramatically change the market’s expectations, I would pick Microsoft.

Not because Microsoft is the most de-risked. It is not. In fact, topological quantum computing probably carries the highest scientific burden of proof of any modality discussed in this post. But that is exactly why it is the breakout contender. If Microsoft’s topological approach works, it does not just add another hardware option to the field. It could change the slope of the entire quantum computing roadmap.

Most quantum computing approaches are fighting the same war: build better physical qubits, reduce error rates, connect more of them together, then spend a large number of physical qubits to create a smaller number of protected logical qubits. That overhead is the tax quantum systems pay for fragility. Superconducting, trapped ion, neutral atom, and photonic systems all have different versions of this problem.

Microsoft is making a more radical bet: what if the qubit itself can be made more naturally resistant to noise? That is the promise of topological quantum computing. In plain English, Microsoft is trying to encode quantum information in a way that is less vulnerable to local disturbances. Instead of treating error correction as something you add later, the ambition is to bake more protection into the physics of the qubit itself. If that works, the system could need fewer physical qubits per logical qubit, making the path to fault-tolerant quantum computing meaningfully more efficient.

This is why Microsoft deserves a different category from the rest of the field. A silicon spin company might surprise the market by scaling faster than expected. A diamond company might surprise the market by finding an early deployment niche. Microsoft could surprise the market by changing the denominator in everyone’s scaling math.

The caveat is important: the community is not fully convinced yet. Microsoft’s Majorana program has been controversial, and outside physicists have raised questions about whether the evidence proves the existence and control of the topological states Microsoft needs. The company’s recent Majorana announcements are promising, but they still need broader independent validation, peer-reviewed evidence, and ultimately system-level demonstrations that show useful qubit operations, not just interesting device physics.

Microsoft does not need to win the NISQ race to matter. It may not compete well on near-term qubit-count comparisons or cloud-access demos. But if it can prove the topological thesis, it could enter the fault-tolerant race from a different angle entirely.

So I would not call Microsoft the safest bet in this category. I would call it the most convex one. The downside case is that the physics remains too hard, too contested, or too slow to commercialize. The upside case is that Microsoft turns out to have been working on one of the few architectures capable of materially reducing the cost of useful fault-tolerant quantum computing.

The technical challenges are easy to name and hard to achieve: scaling, integration, error correction, calibration automation, and system-level reliability. Silicon spin companies need to prove that beautiful device physics can survive wafer-scale variation and large-array control. Donor-atom approaches need to prove that atomic precision can become manufacturing scale. Topological approaches need independent validation of the underlying physics and then actual qubit operations that the community accepts. Diamond systems need to move from deployability to computational relevance.

Signals I’m watching over the next 1–3 years

• Foundry yield data, not just hero devices. I want to see wafer-level statistics, device-to-device variation, and reproducible performance across many chips.

• Public error-correction primitives in silicon. A small logical-qubit or repeated-stabilizer experiment would matter more than another isolated fidelity number.

• Independent access and benchmarks. Cloud, national-lab, or HPC testbeds should publish enough performance data for outsiders to compare systems honestly.

• Cryogenic control progress. Watch for credible reductions in wiring, heat load, control latency, and power consumption without fidelity collapse.

• Semiconductor ecosystem partnerships. Foundry, packaging, metrology, and cryo-electronics partnerships may matter as much, or more, as quantum software partnerships.

• For Microsoft specifically: independent validation. The signal to watch is not another polished topological announcement. It is peer-reviewed, replicated evidence that outside physicists broadly accept.

The broader point is that modality competition is not just a beauty contest between qubits. The winning modality has to plug into software, compilers, error correction, cloud interfaces, networking, standards, customer procurement, and eventually applications.

That is why this “other modalities” category is so interesting. Some of these approaches may fail. Some may become enabling technologies rather than full-stack winners. But if silicon spin, topological qubits, diamond systems, or another solid-state platform breaks through, the market could look very different from today’s view. The quantum stack does not care which physics story we like best. It will reward the platform that can become reliable, scalable, programmable, and economically useful.

That is the real modality race.

Disclosure: The author is a venture investor with investment interests in quantum and may have an interest in companies discussed in this post. The views expressed herein are solely the views of the author and are not necessarily the views of Corporate Fuel Partners or any of its affiliates or any companies it has investment interests in. Views are not intended to provide and should not be relied upon for investment advice.

References:

Castelvecchi, Davide, “Microsoft upgrades controversial quantum chip — researchers are still skeptical,” Nature News, June 3, 2026

Edlbauer, Hermann, et al., “An 11-qubit atom processor in silicon,” Nature, December 17, 2025

Elliott, Francesca, “Diraq Adds Leading Deeptech and Venture Investors to USD $22 Million Investment Round,” Diraq, June 25, 2024

GlobeNewswire, “Diraq Signs $38M Letter of Intent with the U.S. Department of Commerce under CHIPS Act,” GlobeNewswire / Markets Insider, May 21, 2026

Intel Newsroom, “Intel’s New Chip to Advance Silicon Spin Qubit Research for Quantum Computing,” Intel, June 15, 2023

Intel Newsroom “Intel Takes Next Step Toward Building Scalable Silicon-Based Quantum Processors,” Intel, May 2024

Nayak, Chetan, “Majorana 2 – Microsoft’s Scalable Quantum Processor With Reliable, Long-Lasting Qubits,” Microsoft Quantum, June 2026

Quantum Motion, “Quantum Motion Raises £42 Million Investment Round Led By Bosch Ventures,” February 21, 2023; Quantum Motion, “Quantum Motion Delivers the Industry’s First Full-Stack Silicon CMOS Quantum Computer,” September 15, 2025

Turczyn, Coury, “Q&A: Inside Quantum Brilliance’s Quantum Computer Technology,” Oak Ridge Leadership Computing Facility, September 2, 2025.