There is an abundance of evidence that supply chains are undergoing a major transformation, centered around geopolitical alignment. But is this global fragmentation uniform in pattern, and does it extend across every sector? The evidence suggests that the race for domination appears to be more concentrated in national security-sensitive sectors. Semiconductors offer a quintessential case study, with this technology serving as a backbone for national economies and military power alike.
What renders semiconductors prone to weaponisation in this fashion? Merely being an “important technology” does not make this technology a source of geoeconomic power. By observing factors that may make this technology hard to substitute or circumvent, we identify two key structural reasons. First, capital-intensity, and second, technological complexity. With the fulfilment of these two preconditions (combined with the will to weaponize), the structural nature of this technology renders means that it is a potential wellspring of geoeconomic power.
Capital intensity
Advanced semiconductor production requires billions of dollars of investments in R&D and fabrication. A leading-edge plant now costs around $20 billion to build. In critical segments such as EUV lithography (dominated by ASML) and high-performance GPUs (where NVIDIA holds a central position), a small number of firms control the total market share. A disruption at any one of these nodes can impact chipmakers on multiple continents. The more concentrated the market, the greater the strategic exposure should things go wrong.
By employing capital and targeting different stages of semiconductor production, these companies have achieved ascendancy in different areas; amongst each stage of the supply chain, the design phase may be identified as being of proximal importance. Controlling the design phase of the chips is enough for a coercive leverage over the countries that control other stages of production.
Technological complexity
Moore’s Law is slowing down. As the gains from simply shrinking transistors diminish, the industry has shifted toward new chip architectures, advanced packaging, and ever-tighter coordination between designers, fabricators, and equipment makers. This makes supply chains even harder to break apart. Governments are adapting, and what once looked like ordinary commercial risk now looks like a national security vulnerability. The response has been a wave of industrial policy, the CHIPS Act in the United States, similar legislation in Europe and Japan, and a tightening web of export controls on advanced equipment. Firms like ASML and NVIDIA now sit at the intersection of commercial logic and geopolitical imperatives, pulled in both directions at once.
Techno-Nationalism: Securing General-Purpose Technology
Semiconductors are defined as a key component of general-purpose technology: their use ranges from cars and consumer electronics to hospitals, financial systems, and weapons platforms. That is what makes their securitisation so important. When governments treat chip access as a matter of national security, they are extending security logic across the entire productive base. The push to bring supply chains closer to home, accepting higher costs in exchange for less dependence on potentially hostile suppliers presents a structural recalibration.
Chip policy has moved from the trade desk to the national security council, and it is unlikely to go back.
This process has been popularly theorised by what Henry Farrell and Abe Newman termed “weaponized interdependence”: the deliberate use of chokepoints in global economic networks as instruments of coercion. American export controls blocking China’s access to advanced chips and chip-making equipment are the clearest example. But the logic extends beyond any single rivalry. Whoever controls the critical nodes of semiconductor production holds a lever over the economies of rivals. That is why chip policy has moved from the trade desk to the national security council, and why it is unlikely to go back.
