Government

imec

Silicon Spin Private Government Lab Leuven, Belgium
Founded 1984 imec-int.com ↗

Overview

imec is the world's leading independent semiconductor research and development center, headquartered in Leuven, Belgium. Founded in 1984 as a spin-off of KU Leuven, it operates as a non-profit research hub funded by a combination of government grants (primarily from the Flemish government via imec's VIVES program and the broader European research ecosystem) and industrial membership fees from semiconductor companies worldwide. imec's quantum computing program is one of its strategic growth pillars, focused specifically on silicon spin qubits fabricated using advanced CMOS-compatible processes on 300mm wafers — the same tooling and process flows used to manufacture leading-edge conventional chips. This is the core of imec's quantum thesis: that the path to scalable, manufacturable quantum computing runs through the semiconductor foundry, not through bespoke cryogenic hardware developed in physics labs.

imec's commercial strategy in quantum is not to build a quantum computer itself, but to serve as the critical manufacturing and process development partner that enables others to do so. It offers its 300mm CMOS silicon spin qubit platform to quantum startups, academic groups, and large technology companies that need access to world-class semiconductor fabrication without building their own fabs. This positions imec as a kind of neutral infrastructure layer — analogous to what TSMC is to the classical chip industry, though imec's role is more R&D-oriented than high-volume manufacturing. Key partners in the quantum domain include Intel (a long-standing strategic partner across multiple programs), Diraq (the Australian silicon spin qubit startup co-founded by Andrew Dzurak), and various European quantum initiatives under the EU Quantum Flagship program.

The most significant recent milestone is a July 2026 joint demonstration with Diraq of a coherent, readable 8-qubit silicon MOS spin-qubit array fabricated on imec's 300mm CMOS platform. This is a meaningful integration milestone: it combines imec's process technology leadership with Diraq's spin-qubit device architecture at a scale relevant to near-term algorithm development and error correction research. The result validates imec's platform as capable of hosting third-party qubit designs at wafer scale, which is central to its commercial positioning as the quantum foundry of choice for silicon spin qubit developers.

Within the competitive landscape, imec occupies a structurally distinct position from pure-play quantum hardware companies. It is not racing Intel, IBM, or Google directly on qubit count or quantum volume; rather, it is building the process infrastructure upon which silicon spin qubit companies will depend if the modality is to scale. Its defensible position rests on its CMOS process expertise, its installed base of 300mm fab equipment, and its decades-long relationships with the global semiconductor industry. The risk is that this position is inherently enabling rather than proprietary — imec's success depends on silicon spin qubits succeeding as a modality, and on imec remaining the preferred fabrication partner as the field matures.

Leadership

Luc Van den hove
President and CEO

A semiconductor industry veteran who joined imec in 1984 and has led the organization as CEO since 2010, overseeing its expansion into advanced logic, memory, and emerging compute paradigms including quantum.

Wim Dehaene
Fellow and Program Director, Quantum Computing

A KU Leuven professor and imec fellow with deep expertise in VLSI and quantum circuit design, leading imec's quantum computing research program.

An Steegen
Executive Vice President, Semiconductor Technology and Systems

Former TSMC and imec process technology executive responsible for imec's full process R&D portfolio, encompassing both classical and quantum device development.

Jo De Boeck
Chief Strategy Officer

Long-tenured imec executive responsible for strategic partnerships and technology roadmap alignment across imec's research programs.

Technology

imec's quantum technology approach centers on silicon MOS spin qubits fabricated on its 300mm CMOS-compatible process platform. The fundamental thesis is that silicon spin qubits are the only qubit modality intrinsically compatible with semiconductor mass manufacturing, and that leveraging existing CMOS process flows — including lithography, gate stack deposition, and implantation techniques refined over decades — provides a direct path to the qubit counts and uniformity required for fault-tolerant quantum computing. imec does not develop a proprietary qubit architecture per se; instead, it develops the process platform and device layers that third-party qubit designers can build upon, analogous to a PDK (process design kit) model in classical semiconductors.

The July 2026 demonstration with Diraq of an 8-qubit silicon MOS spin-qubit linear array on the 300mm platform is the most specific public data point on current system capabilities. The result demonstrated coherent operation and single-shot readout across all eight qubits, fabricated using imec's CMOS tooling. Specific gate fidelity figures and coherence times for this particular device have not been publicly disclosed in investor-grade detail as of this writing, though silicon MOS spin qubits at this process node typically exhibit single-qubit gate fidelities in the 99–99.9% range and coherence times (T2*) in the microsecond range under optimal conditions — figures that are consistent with but not definitively confirmed for this specific imec/Diraq device. imec also engages with Intel's silicon spin qubit program, which has historically targeted similar device architectures on leading-edge process nodes.

Key technical differentiators for imec include: (1) access to 300mm wafer-scale fabrication, enabling statistical process control and device uniformity at a level inaccessible to university labs or small startups; (2) integration of quantum device layers with CMOS-compatible interconnect and classical control circuitry in the same process flow, critical for eventual cryo-CMOS integration; and (3) a neutral, multi-client model that allows multiple quantum startups and research groups to co-develop on the same platform, accelerating process learning through volume.

Key Systems

Performance Highlights

Financials

imec is a non-profit research institute and does not have publicly traded equity, report earnings in the conventional sense, or raise venture capital rounds. Its financial model is structurally different from commercial quantum hardware companies: it is funded through a combination of structural Flemish government support (via the VIVES strategic research center framework and direct imec funding from the Flemish government, estimated at several hundred million euros annually in aggregate), bilateral research contracts with industrial partners (including membership fees, co-development agreements, and program access fees from companies like Intel, TSMC, Qualcomm, and quantum startups), and European Commission funding through programs such as Horizon Europe and the EU Quantum Flagship. imec's total annual budget is approximately €900 million to €1 billion (figures approximate, based on publicly available annual reports through 2024), with roughly 5,500 staff and researchers.

imec's quantum computing program is one of several strategic pillars within a much larger semiconductor R&D operation; the quantum-specific budget is not separately disclosed. Industrial partners pay access fees to work on imec's quantum platform, which partially offsets the cost of quantum process R&D. There is no revenue in the commercial sense, no burn rate concern in the venture-capital sense, and no IPO trajectory. The financial risk profile is therefore fundamentally different from pure-play quantum startups: imec's quantum program could be reduced or redirected if Flemish government priorities or industrial partner interest shifts, but it is not at risk of the kind of cash runway exhaustion that threatens many quantum hardware companies.

For investors, imec is not a directly investable entity. Its relevance to investment portfolios is indirect: it is a key enabling partner whose platform viability affects the investment cases of companies like Diraq, and its process development work is a bellwether for the manufacturability thesis underlying silicon spin qubit investments broadly.

Key Figures

Milestones

Q3 2026
imec and Diraq demonstrated coherent operation and single-shot readout of an 8-qubit silicon MOS spin-qubit linear array fabricated on imec's 300mm CMOS-compatible platform.

This is the most substantive public validation of imec's silicon spin qubit platform at a multi-qubit scale. It confirms that a third-party qubit architecture (Diraq's) can be successfully implemented on imec's foundry process, establishing imec's multi-client platform model as technically viable. Eight qubits fabricated on 300mm CMOS is a meaningful step toward demonstrating the manufacturability advantage imec claims over bespoke lab-fabricated devices.

2025
imec expanded its quantum computing research program under the EU Quantum Flagship, continuing to serve as a central fabrication partner for European silicon spin qubit research consortia.

Sustained EU Quantum Flagship engagement provides non-dilutive funding for process development and positions imec as the preferred European fabrication hub for silicon spin qubit R&D, reinforcing its role in the broader European quantum sovereignty agenda.

2025
imec continued joint silicon spin qubit process development with Intel, building on a long-running collaboration on CMOS-compatible qubit fabrication at leading-edge process nodes.

Intel's silicon spin qubit program remains one of the best-resourced in the industry. imec's sustained role in that program provides both revenue and process learning that benefits its broader quantum platform.

2024
imec published results on cryo-CMOS control circuit integration research, targeting the co-integration of classical control electronics with silicon spin qubit arrays at cryogenic temperatures.

The wiring bottleneck — delivering control signals to thousands or millions of qubits at millikelvin temperatures — is one of the central engineering challenges for scaled silicon spin quantum computing. imec's cryo-CMOS work addresses this directly and is a prerequisite for any fault-tolerant silicon spin system.

2024
imec formalized its quantum computing platform offering to external clients, making its 300mm silicon spin qubit fabrication process accessible to quantum startups and academic groups through structured co-development agreements.

This move toward a formal multi-client foundry model for quantum is strategically significant: it mirrors imec's proven model in classical semiconductor R&D and creates a recurring revenue stream from the growing cohort of silicon spin qubit startups that lack fab access.

Roadmap

imec's publicly stated quantum computing roadmap focuses on scaling its silicon spin qubit platform toward device counts and process maturity relevant to fault-tolerant quantum computing, with a particular emphasis on manufacturability and co-integration with classical control electronics. The near-term priorities (2025–2027) center on demonstrating two-qubit gate fidelities exceeding 99% on the 300mm platform, improving device-to-device uniformity through process control improvements, and advancing cryo-CMOS integration so that control circuitry can be co-fabricated or closely integrated with the qubit layer. The 8-qubit demonstration with Diraq in mid-2026 represents execution against this near-term phase.

Medium-term roadmap targets (approximately 2027–2030, based on public statements and EU Quantum Flagship milestones) include scaling to tens of qubits per chip with sufficient uniformity for quantum error correction experiments, and demonstrating surface code or similar error correction primitives on the 300mm platform. imec has not publicly committed to specific qubit count targets with precise dates in the way that some commercial quantum hardware companies have, reflecting both its research-institute culture and the genuine uncertainty in silicon spin qubit scaling timelines. Timelines across the silicon spin qubit field broadly have proven optimistic historically, and imec's roadmap should be interpreted with that context.

Longer-term, imec's roadmap envisions its 300mm CMOS platform serving as the fabrication foundation for commercial-scale silicon spin quantum processors developed by its industrial and startup partners — a scenario in which imec functions as the TSMC of quantum silicon. Whether this vision materializes depends on silicon spin qubits demonstrating competitive error correction performance relative to superconducting and trapped-ion competitors, and on imec successfully transitioning from R&D partner to quasi-foundry at meaningful volume. No public timeline has been given for this transition.

Competitive Position

imec's most direct competitors in the silicon spin qubit fabrication space are Intel (which operates its own leading-edge fab and has an internal silicon spin qubit program, making it simultaneously a partner and potential rival), CEA-Leti in France (a comparable European government research institute with its own CMOS-compatible silicon spin qubit program, and the fabrication partner for Quobly and other French quantum startups), and GlobalFoundries or TSMC if either moves to offer quantum-specific process services. Among these, CEA-Leti is the most direct structural analog and competitive threat: it offers a similar R&D-institute-as-foundry model for silicon spin qubits within the European quantum ecosystem, and competes directly with imec for EU Quantum Flagship project participation and for startup partnerships. The Diraq partnership is a meaningful win for imec over CEA-Leti in attracting non-European clients.

imec's defensible advantages are substantial but not permanent. Its 300mm CMOS process maturity and equipment base represent a multi-decade capital and knowledge investment that cannot be replicated quickly. Its semiconductor industry relationships — with Intel, ARM, and dozens of other companies — give it process knowledge and co-development resources that pure-play quantum labs lack. Its neutral, non-commercial status makes it an attractive partner for companies that would be reluctant to share IP with a commercially oriented foundry. However, imec is vulnerable if silicon spin qubits fail to demonstrate competitive error correction performance relative to superconducting qubits (the IBM/Google path) or trapped-ion systems (IonQ, Quantinuum): in that scenario, the entire modality loses investment and imec's quantum program becomes a research footnote rather than a strategic asset. imec is also vulnerable to Intel internalizing more of its quantum process development, reducing imec's role to that of a secondary partner.

Risks & Opportunities

Key Risks

  • Silicon spin qubit modality risk: if superconducting or trapped-ion qubits achieve fault-tolerant operation before silicon spin qubits demonstrate competitive error rates, investment and talent in silicon spin will consolidate around those modalities, marginalizing imec's quantum program regardless of its process capabilities.
  • Intel internalization risk: Intel is simultaneously imec's largest partner and a potential competitor with its own world-class fabs; if Intel's silicon spin program advances sufficiently, Intel could reduce its reliance on imec for quantum process development.
  • CEA-Leti competition for European startup partnerships: CEA-Leti offers a structurally similar model within the EU quantum ecosystem and has strong relationships with French and broader European quantum startups, directly competing for the quantum foundry pipeline imec is building.
  • Cryo-CMOS integration timeline risk: the engineering challenge of integrating classical control electronics with millions of qubits at cryogenic temperatures remains unsolved at scale; delays in cryo-CMOS solutions would bottleneck imec's platform scalability regardless of qubit quality.
  • Government funding dependency: a significant fraction of imec's budget flows from the Flemish regional government and EU programs; shifts in political priorities or EU research funding allocations could constrain the quantum program's resources.
  • Talent competition: leading-edge semiconductor process engineers and quantum physicists are both in short supply globally; imec competes with Intel, TSMC, and well-funded quantum startups for the dual-expertise staff its quantum program requires.

Key Opportunities

  • Silicon spin qubit foundry model: as the cohort of silicon spin qubit startups (Diraq, Equal1, HRL Laboratories, and others) grows, each faces the same problem — they cannot afford their own 300mm fab. imec is uniquely positioned to serve this entire segment as a shared foundry, generating sustained co-development revenue while accelerating its own process learning.
  • EU quantum sovereignty agenda: European policymakers are actively funding domestic quantum hardware capabilities to reduce dependence on US and Chinese quantum technology; imec is the most credible European silicon quantum fabrication hub and stands to receive sustained funding and preferred-partner status in EU-funded quantum programs through the next multi-year budget cycle.
  • Cryo-CMOS integration leadership: if imec can demonstrate viable co-integration of cryo-CMOS control circuits with silicon spin qubit arrays at wafer scale, it would establish a significant technical lead over competitors and create a high-barrier IP position in the qubit control bottleneck.
  • Expansion of the Diraq partnership: the July 2026 8-qubit demonstration is a proof of concept; deepening this collaboration toward 50+ qubit devices and eventually error-corrected logical qubits would generate high-visibility publications, attract additional startup clients, and validate imec's platform at commercially relevant scale.
  • Standards and IP licensing: as silicon spin qubit process recipes mature on imec's platform, there is potential to monetize process IP through licensing to foundries seeking to offer quantum services, analogous to imec's existing model of licensing process know-how to chipmakers.

Investment Considerations

⚑ GroundState Take

The bull case for imec's quantum program — to the extent investors can access it indirectly — rests on three pillars. First, silicon spin qubits have a credible long-term scaling argument that no other qubit modality can match: CMOS compatibility means that if the physics works, the manufacturing infrastructure already exists. Second, imec's structural position as a neutral R&D hub means it benefits from the success of any silicon spin qubit company, not just one; it is in effect a diversified bet across the entire silicon spin ecosystem. Third, the July 2026 Diraq demonstration provides concrete evidence that imec's platform can host third-party architectures at multi-qubit scale, validating the foundry model thesis. For investors in silicon spin qubit companies like Diraq or Equal1, imec's platform maturity is a direct input to those investment cases — a stronger imec platform de-risks the manufacturing execution risk of those companies.

The bear case centers on modality and timing risk. Silicon spin qubits remain behind superconducting qubits on most near-term performance metrics (qubit count, demonstrated error correction), and the timelines for reaching fault-tolerant operation have historically slipped. imec is not directly investable, so even if its quantum program succeeds brilliantly, investors cannot capture that value directly — they can only benefit indirectly through the companies imec enables. Furthermore, imec's non-profit, government-backed structure, while financially stable, also means the quantum program is subject to research-institute decision-making timelines and priorities that may not optimize for commercial outcomes. Investors should view imec primarily as a risk-reduction factor and infrastructure validator for silicon spin qubit portfolio companies, not as a standalone quantum investment.

Recent Digest Coverage

Last updated 2026-10-01 4 digest mentions (past 90 days)