The history of quantum computing is littered with premature proclamations of commercial relevance. The technology has occupied a peculiar liminal space for three decades: too interesting to ignore, too immature to deploy. Every few years, a new milestone - a larger qubit count, a more impressive demonstration circuit, a new error correction scheme - is greeted with breathless coverage suggesting that the quantum era is finally arriving. And every few years, the gulf between laboratory demonstration and commercial utility remains stubbornly wide.
I want to be careful, therefore, not to add to this pattern of premature enthusiasm. But I also want to be honest about what I believe is genuinely different in 2025. After spending the past three years as Lumino Capital's lead for computing investments, including making two quantum computing investments and maintaining close relationships with academic groups at Delft, Oxford, and MIT, I have formed a view that is both more cautious than the broader market optimism and more specific about where commercial opportunity genuinely exists today.
The State of the Art in 2025
The leading quantum computing platforms as of mid-2025 fall into several broad categories, each with distinct characteristics in terms of qubit count, error rates, coherence time, and connectivity. Superconducting qubits - the approach used by IBM, Google, and most start-ups - offer relatively high gate speeds but require extreme cooling to millikelvin temperatures and currently achieve error rates that, while improving, still necessitate substantial error correction overhead. Trapped ion systems - used by IonQ and Quantinuum - achieve lower error rates but slower gate speeds, and face scaling challenges as the ion chain length increases. Photonic systems - the approach that interests us most at Lumino - operate at room temperature and are naturally suited to quantum communication applications, but face challenges in creating the deterministic photon-photon interactions required for universal quantum computation.
The honest summary of the current state is that no platform has yet demonstrated error-corrected logical qubit performance that would enable the fault-tolerant quantum computation required for the canonical quantum algorithms (Shor's algorithm for factoring large integers, Grover's algorithm for unstructured search, quantum simulation of large molecular systems). The qubit counts have grown dramatically - IBM now operates processors with over 1,000 physical qubits - but physical qubit counts are a highly misleading metric in the absence of error correction, since a single logical qubit capable of fault-tolerant computation typically requires hundreds or thousands of physical qubits.
So the era of quantum computers running Shor's algorithm to break current encryption standards, or simulating drug candidate molecules with molecular-scale precision, has not arrived. Anyone claiming otherwise is either confused or deliberately misleading. The question for investors is not whether quantum computing will eventually be transformative - on this, I have no doubt - but where practical commercial value can be extracted from the technology available today and over the next three to five years.
Where Commercial Value Exists Now
The most honest answer is that meaningful commercial quantum advantage exists today in a small number of narrowly defined problem categories. Quantum sensing is probably the most commercially mature application of quantum technology. Quantum sensors - atomic clocks, gravimeters, magnetometers, inertial navigation systems - exploit quantum mechanical phenomena to achieve measurement precision that is fundamentally impossible with classical devices. These are not hypothetical applications; commercial quantum sensors are already deployed in navigation, geophysical surveying, and medical imaging. The companies building the most advanced quantum sensors represent a genuine investment category that is distinct from, and somewhat less risky than, quantum computing per se.
Quantum key distribution, another application of quantum mechanics rather than quantum computation, is also commercially deployed in several high-security communications networks. The underlying physics is sound and the technology is mature. The commercial challenge is integration with existing telecommunications infrastructure, and several start-ups are making meaningful progress on this problem.
For quantum computing specifically, the most credible near-term commercial applications are in optimisation problems - portfolio optimisation, logistics routing, materials simulation - where the quantum advantage is heuristic rather than provable. These applications can extract commercial value from noisy intermediate-scale quantum (NISQ) devices that are far short of the fault-tolerant ideal. The challenge for investors is that the claims being made in this space are difficult to verify, the benchmarks are easily manipulated, and the competitive set of classical optimisation algorithms is much stronger than commonly acknowledged.
The Enabling Technology Opportunity
At Lumino Capital, our investment thesis in quantum computing is somewhat contrarian. Rather than backing companies that are competing to build the best quantum computer - a race that is dominated by well-capitalised incumbents and is likely to remain so - we have focused on companies building enabling technologies that will be essential to any quantum computing system, regardless of which platform ultimately wins.
Cryogenic control electronics is one of the most compelling examples. Current superconducting quantum computers require their qubits to operate at temperatures close to absolute zero, but the classical electronics used to control and read out the qubits operate at room temperature. Connecting these two temperature regimes requires large numbers of coaxial cables that carry noise, consume power, and fundamentally limit the number of qubits that can be operated simultaneously. Companies that can build classical control electronics that operate reliably at cryogenic temperatures - allowing the control hardware to live at the same temperature as the qubits - remove this bottleneck and enable a new generation of much larger quantum processors. This is precisely what ColdSilicon Labs, one of our portfolio companies, is building.
Photonic interconnects represent another enabling technology with a dual-use commercial proposition. Silicon photonics technology that enables high-speed, low-power optical communication between chips has near-term commercial value in data centres, independent of any quantum computing application. The same technology base that we are backing through Photon Link for classical computing applications could, over a longer timeframe, contribute to the development of quantum photonic networks. The near-term commercial case is strong and de-risks the longer-horizon quantum opportunity.
The Investment Framework
Based on our experience, we have developed a framework for evaluating quantum computing investments that we think is more rigorous than most of what circulates in the broader market.
The first filter is scientific credibility. The quantum computing space has attracted a significant number of founders who understand the business opportunity but have limited depth in the underlying physics. We require founding teams to have genuine experimental expertise in the specific qubit modality they are developing - not theoretical understanding, but laboratory experience with actual quantum hardware. The failure modes in quantum systems are highly non-obvious, and teams that lack this experience will encounter them.
The second filter is clarity about the near-term commercial pathway. We are deeply sceptical of quantum computing business plans that rely on applications that will only be commercially relevant once fault-tolerant quantum computers are available. We want to understand how the company will generate revenue over the next three to five years, on hardware that exists or can be built with the funding being raised. If the answer requires assumptions about hardware performance that have not yet been demonstrated, the commercial pathway is not credible.
The third filter is defensibility. Quantum computing is unusual in that the competitive moats are difficult to build without patent protection, since the underlying science is published and the talent pool is concentrated at a small number of research institutions. We look for companies with clear intellectual property positions, typically built around proprietary fabrication processes or materials characterisation techniques that are difficult to reverse-engineer.
Looking Forward
The next five years in quantum computing will, I believe, be characterised by a significant separation between those who achieve demonstrable commercial progress on near-term applications and those who remain in a perpetual state of "the breakthrough is just around the corner." The companies that survive and thrive will be those that have built genuine revenue from the technology that exists today, while continuing to invest in the research that will unlock the transformative applications of the longer-term future.
For investors, the implication is clear: evaluate quantum computing investments on the same rigorous commercial criteria that you would apply to any other deep tech investment. Do not allow the narrative of quantum supremacy to substitute for a coherent near-term business model. And do not confuse physical qubit count - the metric that dominates press coverage - with commercial readiness. The most exciting quantum companies we are currently reviewing are not the ones with the most impressive qubit numbers, but the ones that have identified specific industrial problems where the quantum advantage is real, measurable, and commercially valuable today.