Quantum Sovereignty and the Utility Trap

This article is part of my Quantum Utility Map Deep Dive series. It builds on the technical research in The Quantum Utility Ladder and the competitive analysis in Quantum Computing by 2033, examining what the concentration of quantum advantage in specific industries means for national strategy and supply chain resilience.

Technology chokepoints and the quantum utility trap

In 2019, Huawei learned what supply chain dependency means when Google, complying with U.S. export rules, suspended its license to Google’s Android apps and services. In 2022, TSMC stopped shipping advanced chips to Russian customers within days of sanctions being imposed. In 2023, the Netherlands began requiring export licenses for ASML’s most advanced immersion lithography systems. In each case, a technology that had been treated as a commercial commodity was revealed to be a strategic chokepoint.

Quantum computing is heading toward the same inflection point, but with a complication that makes the stakes higher. My analysis of fault-tolerant quantum algorithms shows that quantum advantage is concentrated in a small number of industries: pharmaceutical R&D, chemical catalyst design, battery chemistry, and advanced materials science. These are not peripheral sectors. They are the industries that determine whether a nation can develop its own drugs, design its own energy systems, and manufacture its own advanced materials.

This creates what I call the utility trap: the more useful quantum computing becomes for your critical industries, the more dangerous your dependency on whoever controls the quantum supply chain. Utility and vulnerability scale together.

Where the trap closes

At 200 logical qubits (projected around 2029), quantum computing begins producing scientific results that feed into industrial R&D workflows. Condensed-matter simulations, small catalytic fragments, early materials modeling. The dependency at this stage is manageable. Classical alternatives exist for most problems, even if they are slower or less accurate. Losing quantum access would slow your research; it would not stop it.

At 2,000 logical qubits (projected around 2033), the picture changes. As I detailed in the competitive analysis, quantum computing at this scale can simulate the enzyme mechanisms of drug metabolism, the catalyst active sites that control industrial chemical efficiency, and the battery degradation pathways that limit next-generation energy storage performance. Your competitor with quantum access can answer molecular questions that your classical computers cannot answer at any speed.

At 10,000+ logical qubits (late 2030s), quantum computing stops being an edge in these sectors and becomes a requirement. Full-scale materials modeling, complete catalytic cycle simulation, protein-ligand binding at pharmaceutical accuracy. An organization or nation without access to quantum computing at this scale cannot compete in pharma R&D, advanced materials design, or next-generation energy technology.

The trap is that each step up the utility ladder tightens the dependency. By the time quantum computing is essential, the supply chain relationships, the vendor contracts, the architectural choices, and the expertise gaps that determine whether your access is sovereign or dependent have already been locked in. The decisions that matter are made years before the consequences become visible.

Dependencies below the QPU

Most discussions of quantum computing access focus on the question of cloud versus on-premises deployment. This framing is dangerously incomplete.

A nation that purchases an on-premises quantum computer from a foreign vendor has acquired hardware. Whether it has acquired capability depends on a chain of dependencies that extends far below the level of the QPU itself.

A superconducting quantum computer depends on six things below the QPU:

• A dilution refrigerator, cooled to approximately 15 millikelvins (colder than deep space), manufactured by a small number of companies, with Finland’s Bluefors holding dominant market share.
• Helium-3, the isotope used in the cooling mixture, recovered mainly from the decay of tritium in U.S. and other nuclear programs and controlled for export from the United States under ECCN 1C232. It is extraordinarily expensive, and I have taken the per-kilogram figure out of this article because I could not source it to a dated quote. The mixture in a working refrigerator is largely recirculated, so the constraint is on acquiring an initial charge and replacing losses, not on recurring fuel. That chokepoint has exits, and I have assessed all of them: producing more of the gas, needing less of it, or replacing the millikelvin stage altogether. In its last helium-3 shortage, the United States cut demand by redesigning the radiation portal monitors that were consuming the gas, rather than by finding more of it.
• Specialized microwave control electronics, concentrated in companies based in the Netherlands, Israel, and Switzerland. The most consequential engineering answer to that chokepoint arrived in 2026 and promptly changed hands: HRL Laboratories built a silicon quantum processor that generates all its own time-varying control signals inside the cryostat, pairing an 18-qubit exchange-only spin chip at millikelvin with a roughly 70-million-transistor CMOS controller at 4 kelvin over a 296-line superconducting ribbon cable. IBM announced a definitive agreement to acquire HRL from Boeing and General Motors on July 23, 2026; the Nature paper appeared on July 29; the acquisition completed on August 26. A chokepoint with an engineering solution is still a chokepoint if one company buys the lab that solved it.
• Calibration expertise held by a small number of engineers worldwide.
• Error correction IP embedded in proprietary firmware.
• Ongoing maintenance contracts that can be terminated.

An on-prem IBM system with a service contract that can be revoked is no more sovereign than a cloud API. The dependency has merely changed form; it has not been eliminated.

As I argue in Quantum Sovereignty, this is a feature of the quantum supply chain that distinguishes it from almost every other technology domain. The semiconductor supply chain, for all its concentration at certain nodes (TSMC for advanced fabrication, ASML for lithography), is a massive, multi-hundred-billion-dollar industry with deep redundancy at most stages. Dilution refrigerators, quantum control electronics and isotopically pure silicon are each supplied by a handful of firms at volumes and revenues far below semiconductor scale. The unit counts and revenue figures this article used to give were my own estimates, and I have taken them out until I can attach dated capacity or financial data to them.

The quantum supply chain is thin at almost every node. A fire at a manufacturing facility, a corporate bankruptcy, a shortage of helium-3, a new export control rule: any of these events could set back quantum computing programs across multiple countries for months or years.

When I first wrote that paragraph, I based it on vendor concentration and inference. Six industry associations have since measured it. The Quantum Economic Development Consortium published Global Quantum Supply Chain Flows in September 2026 with QuIC in Europe, Quantum Industry Canada, UKQuantum, Japan’s Q-STAR, and the Korea Quantum Industry Association. Of 160 companies headquartered in 15 countries, 90% named at least one supplier outside their home country and 74% at least one foreign customer, on different denominators of 146 and 133 respondents. They named suppliers in 36 countries and customers in 49. Of the 146 companies that answered the supplier question, 37 named China as a supplier location; nine named it as a customer location. The data were collected in March and April 2026 and published in September. The survey measures cross-border dependence. Vendor concentration, substitutability and disruption risk are outside its scope. At the rates it found, a domestic quantum computer, in this industry, means a machine assembled at home from parts bought abroad.

The other half of the same question is who invented each layer, not who ships it. I traced the inventors and suppliers of every layer of the US stack in Every US Quantum Computer Runs on Foreign Parts and Foreign Inventions.

Hardware choice is supply chain strategy

One of the least appreciated dimensions of the quantum hardware diversity that I have written about extensively is its supply chain implications. Each quantum computing modality depends on a different set of components and suppliers, and for nations seeking sovereign optionality, the choice of hardware approach is, in part, a supply chain choice.

Superconducting quantum computing (IBM, Google) has the most concentrated and most vulnerable supply chain. It requires dilution refrigerators, helium-3, cleanroom fabrication, and specialized control electronics from a small number of foreign suppliers. A nation that commits exclusively to superconducting quantum computing commits to a supply chain with multiple foreign chokepoints. IBM is redesigning the layer around one of them: in August 2026 it connected and cooled its first pair of modular cryogenic cells at Poughkeepsie, box-shaped aluminum cells in place of the familiar cylindrical cryostats, reaching 4 kelvin in under five days and then below 15 millikelvin. IBM designed the enclosure, shielding, wiring and integration architecture, and briefing coverage reports that the cooling technology inside comes from Bluefors. The refrigeration dependency remains. What changes is that the cryostat is now built around IBM processors and IBM couplers, which makes IBM’s own machines easier to scale and makes it harder for a customer to combine components from different vendors.

Photonic quantum computing (PsiQuantum, Xanadu) has a supply chain that overlaps with the existing semiconductor and telecommunications industries. Photonic qubits can be fabricated in modified semiconductor foundries. The cryogenic requirement is real but milder: PsiQuantum describes cooling its chips with liquid helium through a cryogenically cooled copper backplane, and states that its systems operate at temperatures more than a hundred times higher than superconducting qubits require. That is a meaningful difference from a millikelvin dilution refrigerator and it is not room-temperature operation, which this article previously claimed. The optical components they require (lasers, detectors, waveguides, fiber optics) are produced by a large and geographically diverse telecommunications industry. This is part of the strategic logic behind Australia’s major bet on PsiQuantum: photonic quantum computing aligns with supply chain strengths that Australia can realistically develop and control.

Trapped-ion and neutral-atom approaches (Quantinuum, IonQ, QuEra, Pasqal) fall between these extremes. They require precision laser systems and ultra-high-vacuum technology, but not dilution refrigerators or helium-3. IonQ’s recent photonic interconnect demonstration linking two commercial systems is relevant here: modular architectures connected by photonic links offer a path to scaling that distributes the supply chain across optical networking and precision instrumentation rather than concentrating it in cryogenic infrastructure. France’s investment in Pasqal’s neutral-atom approach and the UK’s trapped-ion emphasis through Quantinuum reflect, among other considerations, supply chain calculations about which dependencies are most manageable for European nations.

Silicon spin qubits represent yet another supply chain profile. A comprehensive assessment published in Nature Reviews examined the compatibility between silicon spin qubits and existing CMOS semiconductor fabrication. If spin qubits can be manufactured in conventional or lightly modified semiconductor fabs, the supply chain implications are profound: the existing global semiconductor manufacturing base (TSMC, Samsung, Intel, GlobalFoundries) becomes the quantum hardware supply chain. This is the ultimate sovereign optionality play for any nation with domestic semiconductor fabrication capability. The technical challenges remain significant, particularly in achieving the qubit density and control precision needed at scale, but the manufacturing alignment is unmatched by any other modality. Through 2026 that alignment started producing devices and not only papers: Diraq demonstrated an eight-qubit array fabricated in a commercial foundry, and HRL’s cryo-CMOS result above is a silicon-spin result. I set five tests for the modality in March and marked my own scorecard in August, including the one I had written too weakly.

The strategic implication: hardware modality selection is not only a technical decision about which approach will achieve fault tolerance first. It is a sovereignty decision about which dependencies a nation is willing to accept, which chokepoints it can mitigate, and which foreign relationships it must maintain.

The semiconductor precedent

The political scientists Abraham Newman and Henry Farrell have described the strategic exploitation of network dependencies as the “weaponization of interdependence”: using chokepoints in global economic networks as instruments of coercion. Their framework, developed for financial networks and internet infrastructure, applies with particular force to quantum computing, where the chokepoints are physical and the concentration is, if anything, greater.

We do not need to theorize about how this plays out. The semiconductor industry has provided a detailed preview.

ASML has been unable to ship its most advanced EUV lithography systems to China since 2019, following sustained U.S. diplomatic pressure on the Dutch government. In October 2022, the United States imposed sweeping export controls on advanced AI chips and chipmaking equipment. The Dutch government followed with license requirements for advanced DUV immersion systems in 2023 and extended them in September 2024 to additional DUV systems and to specially designed parts, software and technology. A license requirement is not a blanket ban on servicing every installed machine. American EDA software companies (Synopsys, Cadence, Siemens EDA) restricted Chinese access to chip design tools. NVIDIA was barred from selling its most powerful AI accelerators.

The effects were immediate and severe. Chinese semiconductor programs were disrupted. But the medium-term effects are more instructive: China accelerated domestic substitution programs across every restricted category. Huawei developed its own chip design capabilities. SMIC pushed the boundaries of older lithography nodes. An entire domestic supply chain received urgent investment.

The paradox of technology denial is that restrictions create short-term disruption but long-term self-sufficiency in the target nation. Export controls are a depreciating asset whose strategic value diminishes with each year the target invests in alternatives.

For quantum computing, the export control regime is already in motion. The U.S. Bureau of Industry and Security (BIS) September 2024 rule imposed worldwide controls on quantum computing items, including computers, components (dilution refrigerators, cryogenic circuits), and related software. The Entity List includes Chinese quantum entities, among them CAS institutes, USTC and QuantumCTek; one May 2024 action alone added 22 for quantum-related activities. I have dropped the running total this article used to give, because the list changes and a stale count is worse than none. Outbound investment controls restrict what American investors can put into Chinese quantum companies.

These controls have the same structural characteristics as the semiconductor restrictions: immediate disruption for the target, accelerated domestic substitution, and the creation of parallel supply chains that will eventually reduce Western bargaining power.

But here is the part that most analyses miss: even nations within the allied coalition face significant supply chain dependencies on each other. A European nation that relies on Finnish dilution refrigerators, Dutch control electronics, and American cloud platforms for its quantum computing capability is dependent on the continued goodwill and strategic alignment of three allied governments. Alliances are not permanent, and interests are not always aligned. The history of technology governance is full of examples of allied nations disagreeing about export control policy and technology transfer terms.

What sovereignty looks like in practice

For the vast majority of nations, full supply chain autarky in quantum computing, with every component built domestically, is neither feasible nor desirable. What sovereignty requires is optionality: the ability to pivot when circumstances change.

As I describe in Quantum Sovereignty, that optionality requires three capabilities.

First, systems integration expertise. A cadre of domestic engineers who understand the entire quantum computing stack and can assemble, calibrate, operate, and maintain quantum systems from components sourced from diverse vendors. This is the most cost-effective investment a nation can make. Without it, even a nation with access to the best components in the world will be dependent on foreign integrators to make those components work.

Second, domestically hosted infrastructure. Quantum computing systems installed on the nation’s own soil, under its legal jurisdiction, operated by its own personnel. Cloud access to foreign quantum platforms is valuable for research and education, but any computation involving classified data, sensitive intellectual property, or national security applications must be performed on infrastructure the nation controls.

Third, and most critically, architectural choice. The deliberate adoption of open, modular architectures that preserve the ability to swap components, change vendors, and adapt to new technologies as they emerge. A nation that builds its quantum infrastructure on proprietary, vertically integrated platforms has chosen lock-in. A nation that builds on open architectures has chosen freedom. This choice is made at the procurement stage, and once made, it is extraordinarily expensive to reverse.

Geopolitics is becoming architecture: the design choices embedded in quantum systems (open or proprietary interfaces, modular or integrated components, one supplier or several) determine whether sovereign optionality is preserved or foreclosed.

Real-world examples already exist. The Netherlands’ Tuna-5 is a Dutch-led integrated system hosted on the Quantum Inspire platform operated by QuTech. The QPU is a QuantWare superconducting processor. The control electronics are from Qblox. Systems integration was performed by Dutch engineers who understand every layer of the stack. Tuna-5 is modest in qubit count. What it represents is a nation that can build, operate, maintain, and upgrade its own quantum computing infrastructure without depending on any single foreign vendor.

Israel’s Quantum Computing Center (IQCC) takes a different approach: a QuantWare QPU integrated with Quantum Machines control electronics, combining Dutch processor technology with Israeli control systems expertise. Both examples illustrate the same principle: sovereignty comes from integration capability across components, not from manufacturing every component yourself.

That principle has a hard prerequisite, which is that somebody will sell you the processor. I applied a purchase-order test to the whole field in The Merchant QPU Market: not cloud credits, not a turnkey machine, but a chip in a crate with a price on it. Six companies passed when I applied it, on my own classification and at that date. The clearest example of the integration model in practice is Rigetti’s March 2026 sale of a 9-qubit Novera chip to the University of Saskatchewan, a unit publicly listed at $900,000, destined for a dilution refrigerator from Zero Point Cryogenics in Edmonton, driven by control electronics from Qblox in Delft, and tuned with QuantrolOx software written in Espoo. Four companies, four countries, and the responsibility for the finished computer sitting with the buyer. I run a firm that assembles machines from this supply chain and I have argued the component-industry thesis at book length, so read my enthusiasm with that interest declared.

Whether that market stays open is a policy question. The European Quantum Industry Consortium has pointed out that the Commission’s Chips Act 2.0 proposal gives quantum fabrication no route to strategic-project status, the instrument the proposal uses to back advanced capacity, while keeping the quantum pilot lines. A continent can fund pilot lines and still have nowhere to buy a processor.

What organizations and nations should do

The quantum supply chain will grow more concentrated, more contested, and more subject to geopolitical friction over the coming decade. Organizations and nations that depend on quantum computing for critical industries need to act before the dependencies are locked in.

Map your dependencies. Every quantum program should maintain a detailed inventory of its supply chain dependencies: which components come from which suppliers and where the materials inside those components come from. It should also flag every single-source chokepoint and every item subject to export control restrictions. This mapping must extend below the tier-one supplier level to the tier-two and tier-three suppliers.

Build multi-vendor optionality. Where multiple suppliers exist for a given component, procurement strategies should deliberately distribute orders across them, even at some cost premium, to avoid creating single-vendor dependency. Where only a single supplier exists, actively encourage the development of second sources through trusted international partnerships.

Invest in local expertise. The ability to calibrate, maintain, and troubleshoot quantum systems without relying on the vendor’s engineering team is the minimum threshold for operational sovereignty. This requires training programs, institutional knowledge retention, and long-term investment in quantum engineering education.

Specify open architectures in procurement. Every quantum computing procurement contract should include provisions for open interfaces, data portability, component interchangeability, and vendor exit rights. A turnkey system from a prestigious foreign vendor that cannot be maintained, upgraded, or replaced without that vendor’s cooperation has not purchased capability. It has purchased dependency.

Consider hardware modality as a strategic variable. For nations with semiconductor fabrication capability, photonic quantum computing offers a supply chain that aligns with existing industrial strengths. For nations with precision instrumentation expertise, trapped-ion or neutral-atom approaches may offer more manageable dependency profiles. The choice of modality shapes the sovereignty equation for decades.

The view from the trap

The quantum utility trap has a specific structure. The industries where quantum computing creates the most value (pharma, chemicals, batteries, materials) are the industries most critical to national economic security and technological self-determination. The hardware required to serve those industries is concentrated in a small number of companies and countries. And the architectural decisions that determine whether access is sovereign or dependent are being made now, years before the full consequences become visible.

Nations and organizations that recognize this structure have a window to build optionality: diversified vendor relationships, local integration expertise, open architectures, and strategic supply chain reserves. Those that treat quantum computing as just another cloud service to be procured on the open market will discover, at the worst possible moment, that the market is neither open nor reliable.

The semiconductor industry already taught this lesson. Buyers of quantum computers have to apply it at procurement, before the dependencies are locked in.

For the technical analysis of which industries quantum computing will transform, see The Quantum Utility Ladder. For the competitive implications by industry, see Quantum Computing by 2033. For the complete framework for national quantum sovereignty assessment, see Quantum Sovereignty: Strategic Leadership in the Quantum Era.

From the author