Few sectors carry the strategic weight or exhibit the level of concentration seen in semiconductor manufacturing. Today’s chips underpin everything from artificial intelligence and high-speed finance to defense technologies and industrial automation.Yet the tools needed to produce the most advanced processors come from a single company. ASML in the Netherlands controls more than 90 percent of the global market for extreme-ultraviolet lithography systems. Each machine costs more than 300 million euros, weighs around 180 tonnes, and requires thousands of specialists for assembly.
The manufacturing end of the chain is similarly concentrated. TSMC in Taiwan produces the large majority of the world’s leading-edge chips, while Samsung and SK Hynix operate the remaining capacity. This dependence has become a geopolitical vulnerability. Against this background, a start-up from California is attempting something many in the industry once considered impossible: rewriting the core process of chip manufacturing.
Substrate, founded three years ago, wants to replace EUV lithography with a system based on X-ray light generated by compact particle accelerators. The company has raised more than USD one hundred million from backers including Peter Thiel and In-Q-Tel, the investment arm associated with the US intelligence community. Its valuation stands at around one billion US dollars, despite having no commercial tool. For the United States, the project aligns closely with national efforts to rebuild domestic chip manufacturing. For the wider industry, it represents a potential alternative at a time when global demand is rising sharply.
A Global Industry Reaching the Limits of Concentration
The semiconductor market generated around 600 billion USD in revenue in 2024 and is expected to grow to one trillion USD by 2030. Yet only a few companies are capable of producing chips below seven nanometres. TSMC alone accounts for more than 90 percent of global capacity for the most advanced nodes. Firms such as Nvidia, AMD, Qualcomm and Broadcom rely almost entirely on TSMC for their latest processors.
The capital intensity is enormous. A modern leading-edge fabrication plant requires between 15 and 25 billion USD in initial investment. After construction, further spending is required for staff, process engineering and yield improvement. The complexity of EUV has created a barrier to entry that even governments struggle to overcome.
The United States has committed 52 billion dollars in subsidies through the CHIPS Act. TSMC is spending more than 40 billion dollars on its Arizona sites. Intel’s long-term expansion plan across several states exceeds one hundred billion American dollars. Despite these sums, capacity for leading-edge production is growing only slowly.
Substrate’s Attempt to Replace EUV
Substrate proposes using intense X-ray beams produced by accelerating electrons to near-light speed and guiding them through magnetic structures known as undulators. The radiation that emerges is extremely bright, less prone to scattering than ultraviolet light and capable of producing sharp patterns on silicon wafers.
Analysts at Yole estimate that, if the technology matures, the cost of producing a wafer could fall from around 100’000 to 10’000 euros. Lower capital requirements would open the field to new manufacturers and reduce global dependency on a small number of factories.
The physics underpinning the idea is established. X-rays can penetrate deeper and maintain tighter beam coherence, which reduces the need for the elaborate optical correction stacks required in EUV systems. Substrate claims that X-ray lithography could become a single-exposure process, which would simplify tool design and potentially improve throughput.
Interest in alternative approaches is growing. Inversion Semiconductor is developing table-top particle accelerators, while XLight in California is working on free-electron lasers for chip production. ASML explored accelerator-based light sources in earlier research programmes and owns patents in the field, although it ultimately focused on laser-generated EUV.
Technical and Industrial Challenges
Turning the idea into a production platform remains a major engineering task. High-intensity X-rays require fabrication under ultra-high vacuum conditions. New photoresists must be developed to tolerate the radiation levels. Process-control systems would need to be redesigned from the ground up. A single accelerator may support several lithography tools, which means that any malfunction could halt an entire line.
Analysts at Jefferies estimate that, under favourable conditions, commercial maturity could be a decade away. The benchmark to match is high. Today’s EUV machines deliver more than 200 wafers per hour with high uptime. Substrate aims to begin series production in 2028, but many in the industry see the mid-2030s as the earliest realistic timeframe.
A Project With Strategic Backing
Despite the engineering uncertainty, the political interest surrounding Substrate is significant. Senior officials from the US Department of Commerce and the Department of Energy have visited the company. In-Q-Tel’s investment signals alignment with broader national-security considerations.
The United States seeks greater autonomy over the production of advanced chips, which are critical for defence, infrastructure and future industrial applications. More than 90 percent of the world’s leading-edge AI chips are produced in Taiwan. A technology that could reduce the cost of domestic production is therefore highly attractive. Substrate benefits from a long tradition of public-private collaboration in the United States, where early government engagement often accelerates deep-tech development.
Why This Matters for Switzerland
For Switzerland, developments in international semiconductor manufacturing are more than a distant industrial story. The country invests around 24 billion CHF per year in research and development, equating to roughly 3.4 percent of GDP. Its universities and applied-research centres hold deep expertise in areas directly relevant to next-generation chip technologies.
ETH Zürich and EPFL generate between 25 and 30 deep-tech spin-offs each year. The Swiss Centre for Electronics and Microtechnology (CSEM) in Neuchâtel, with a budget of around 110 million CHF, supports national and international partners in microfabrication, nanolithography, photonics and metrology.
Switzerland also plays a significant role in the industrial value chain. Companies such as STMicroelectronics employ around 1000 people across Geneva and Neuchâtel. Firms including Sensirion, u-blox and LEM supply sensors, connectivity chips and power electronics to global markets. The Swiss photonics industry generates more than 5 billion Swiss francs in exports annually and forms one of the world’s densest clusters for high-precision optical engineering.
In this context, the emergence of alternative lithography methods is highly relevant. If capital costs decline and new production paradigms become viable, parts of the semiconductor industry may become more accessible to countries that specialise in high-precision subsystems and research-driven innovation. Switzerland fits this profile well.
Opportunities for Swiss Deep-Tech Companies
Substrate’s approach offers a useful perspective for Switzerland. The country does not need to build large-scale chip factories to benefit from global shifts. Instead, it can reinforce its role as a provider of enabling technologies.
Swiss companies already excel in optical components, vacuum systems, laser technology, positioning equipment and measurement tools. These are all fields that would be essential for any future X-ray lithography platform. ETH spin-offs contribute to quantum systems, integrated photonics and novel materials. CSEM’s microfabrication expertise could become valuable if alternative lithography methods require new process modules.
Switzerland also has frameworks that support early-stage deep-tech development, including Innosuisse flagship programmes, national quantum initiatives and armasuisse innovation partnerships. These programmes could be expanded to ensure that Swiss firms gain early exposure to new international research networks and prototype platforms.
A More Connected Innovation Landscape
The semiconductor race is entering a more experimental phase. As demand for advanced computing accelerates, new light sources, new materials and new process architectures are being explored globally. Switzerland is well positioned to contribute because it combines long-term research investment with industrial precision.
If new lithography technologies achieve technical and economic viability, new supply-chain structures will emerge. This could create opportunities for countries that specialise in high-value components rather than mass production. Switzerland’s established strengths in precision measurement, optics and automation could become even more relevant.
A Global Competition Switzerland Can Help Shape
The future of chip manufacturing will not depend on a single technology. EUV will remain the industry standard for years, but alternative approaches are gaining traction. As the semiconductor market heads towards USD one trillion, capacity and optionality will matter.
Substrate’s attempt to rethink lithography highlights that even highly consolidated industries are open to disruption when scientific ambition, industrial engineering and strategic support converge. Whether the company ultimately succeeds or not, the project signals a shift in how the next generation of semiconductor technologies may develop.
For Switzerland, this represents an opportunity rather than a challenge. The country’s strengths in research and engineering allow it to play a formative role in the emerging landscape of advanced manufacturing. As new technologies reach maturity, Switzerland can help shape the tools and components that make future chip production possible.
References (APA)
- ASML. (2024). Annual Report 2023. Veldhoven, NL: ASML Holding. https://www.asml.com
- CSEM. (2024). Annual Report 2023. Swiss Centre for Electronics and Microtechnology. https://www.csem.ch
- European Commission. (2023). EU Chips Act: Regulation (EU) 2023/1781. Publications Office of the European Union.
- Gartner. (2024). Forecast Analysis: Semiconductors, Worldwide, 2024–2030. Gartner Research.
- In-Q-Tel. (2024). IQT Portfolio Overview. In-Q-Tel, Inc. https://www.iqt.org
- Jefferies Group. (2024). Semiconductor Capital Equipment: Industry Outlook 2024–2035. Jefferies Equity Research.
- McKinsey & Company. (2023). The Semiconductor Decade: A Trillion-Dollar Industry. McKinsey Global Institute.
- Sensirion Holding AG. (2024). Annual Report 2023. Staefa: Sensirion.
- STMicroelectronics. (2024). Company Profile & Financials 2023. Geneva: STMicroelectronics N.V.
- Swiss Federal Statistical Office (FSO). (2023). Research and Development in Switzerland 2021–2023. Neuchâtel: FSO.
- Swiss National Science Foundation (SNSF). (2023). Research Funding Statistics 2022. Bern: SNSF.
- Swiss Photonics Cluster. (2023). Photonics Industry Report Switzerland 2023. Swissmem / Swiss Photonics.
- Swiss Venture Capital Report. (2024). Venture Capital Report 2023. Startupticker.ch & SEBA.
- TSMC. (2024). Corporate Factbook 2023. Taiwan Semiconductor Manufacturing Company. https://www.tsmc.com
- U.S. Department of Commerce. (2024). CHIPS for America: Funding and Strategy Document. Washington, DC.
- U.S. Department of Energy. (2023). High-Performance Semiconductor Manufacturing: Technical Opportunities. DOE Office of Science.
- Yole Intelligence. (2024). Lithography Equipment and Materials 2024 Report. Yole Group.
