The Semiconductor Sovereignty Race: Why Every Major Economy Is Funding Its Own Chip Industry
The global semiconductor industry is undergoing the most dramatic geographic restructuring in its 70-year history. What was optimised over three decades into the most efficient globally distributed supply chain ever assembled — design in the US, equipment from the Netherlands and Japan, wafers from Taiwan and South Korea, packaging from Malaysia and Thailand, final integration in China — is being deliberately dismantled by governments that have concluded that chip dependency is a strategic vulnerability they can no longer accept. The result is simultaneous, large-scale public investment in semiconductor manufacturing capacity across the United States, European Union, Japan, India, and South Korea, each building or subsidising domestic fabs for reasons that are geopolitical rather than economic. The semiconductor sovereignty race is not primarily a story about technology — it is a story about power, and the unwillingness of major economies to depend on adversaries or unstable partners for the hardware that runs everything from weapons systems to payment networks.
How the Dependency Was Built and Why It Became Unacceptable
The geographic concentration of advanced semiconductor manufacturing was not an accident — it was the outcome of deliberate capital allocation decisions across decades. Taiwan Semiconductor Manufacturing Company's dominance of leading-edge logic chip manufacturing — TSMC produces approximately 90% of the world's chips at 7 nanometers and below — reflects 40 years of compounding investment in process technology, workforce, and supply chain that no competitor has matched. Samsung's position in DRAM and leading-edge logic, and South Korea's dominance of NAND flash memory through SK Hynix, similarly reflect decades of sustained capital commitment that created technical and manufacturing advantages with enormous barriers to replication. The concentration was tolerated because Taiwan and South Korea were stable, aligned with Western security interests, and produced chips at costs that no domestic alternative could approach. Russia's full-scale invasion of Ukraine in February 2022 changed the calculation. The scenario that had seemed hypothetical — a major geopolitical disruption severing access to a concentrated critical supply chain — became observable reality for European natural gas, and the inference for semiconductor supply chains was immediate and obvious.
The Biden administration's CHIPS and Science Act, signed in August 2022 and allocating USD 52.7 billion in semiconductor subsidies, was the legislative crystallisation of the new strategic consensus. The EU Chips Act, adopted in September 2023 with EUR 43 billion in public and private investment targets, followed. Japan's semiconductor investment programme — funding TSMC's Kumamoto fab, Rapidus's leading-edge ambitions, and domestic equipment company expansion — has committed over JPY 4 trillion. India's Semiconductor Mission is subsidising fabs from Micron, Tata Electronics, and CG Power. The scale of simultaneous public intervention in a single industry is unprecedented in the history of industrial policy, and it is occurring precisely because market forces alone would never produce the geographic diversification that security planners require.
What Is Actually Being Built — and Where
The United States is attracting the most ambitious manufacturing investments. TSMC's Arizona complex — two N4 fabs operational or under construction and an N2 fab planned — represents over USD 65 billion in committed capital and the first leading-edge logic manufacturing capacity on US soil in over two decades. Intel's Ohio fab complex, targeting Intel 18A process technology and positioning Intel as a US-based foundry competitor to TSMC, represents USD 28 billion in initial commitment with USD 100 billion planned over the decade. Samsung's Taylor, Texas facility adds another USD 17 billion in committed advanced logic manufacturing. The constraint is not capital or corporate commitment — it is workforce. The US semiconductor workforce gap is estimated at 67,000 skilled workers over the next decade, a shortage that no university pipeline can fill on the timeline that fab construction is demanding. The CHIPS Act's workforce development provisions are addressing this, but trained process engineers and technicians take years to develop.
Europe's ambitions are more modest and more contested. Intel's EUR 30 billion Magdeburg fab — Europe's most significant semiconductor manufacturing commitment — is now delayed, with Intel citing market conditions and cost overruns that have produced a public subsidy negotiation that illustrates the difficulty of sustaining political commitment to 10-year capital projects across electoral cycles. TSMC's Dresden facility, a joint venture with Infineon, NXP, and Bosch targeting automotive and industrial chips rather than leading-edge logic, is advancing more smoothly precisely because it targets mature process nodes where European demand is concentrated. Japan's Rapidus project — aiming to produce 2nm chips domestically by 2027 in partnership with IBM — is the most technically ambitious sovereign chip project outside the US-Taiwan duopoly, and also the most uncertain, given that no Japanese company has manufactured leading-edge logic since Toshiba's memory business was sold in 2018.
The Export Controls That Are Accelerating the Race
The US Commerce Department's October 2022 semiconductor export controls — restricting the sale of advanced chips, chipmaking equipment, and related software to China — and their subsequent expansions in 2023 and 2024 have transformed the semiconductor sovereignty race from a defensive posture into an active technology contest with China. The controls target three categories: advanced logic chips (above certain performance thresholds), high-bandwidth memory, and the semiconductor manufacturing equipment from Applied Materials, Lam Research, KLA, ASML, and their Japanese equivalents that produces them. The effect has been to accelerate Chinese domestic semiconductor development — SMIC's 7nm production using deep ultraviolet lithography rather than EUV, Huawei's Kirin 9000S chip, and China's massive investment in mature process node capacity — while preventing China from accessing the leading-edge equipment needed to close the gap in the near term. The controls have also forced ASML, the Dutch monopolist in EUV lithography, to cease shipments to China entirely — a decision with profound implications for Dutch industrial policy and for China's 2nm ambitions.
The Economics That Make Sovereign Chips Expensive
The uncomfortable truth that proponents of semiconductor sovereignty rarely foreground is that the chips being produced in politically motivated fab investments will cost significantly more than equivalent chips produced in Taiwan or South Korea. TSMC's Arizona fabs are estimated to produce chips at 25–50% higher cost than equivalent Taiwan production, reflecting US labour costs, supply chain immaturity, and the overhead of operating in a regulatory environment designed for safety rather than manufacturing velocity. Intel's 18A process — even if it achieves the yield rates required for commercial viability — will face the same cost premium. The strategic argument is that this premium is insurance: the cost of supply chain resilience rather than the cost of manufacturing efficiency. The market test will come when the subsidies expire and governments must decide whether to perpetuate support for manufacturing that cannot compete on cost with Asian alternatives. The history of industrial policy suggests this decision is harder than the initial investment, and that the sovereign chip projects are committing their governments to sustained support whose long-term cost is substantially underestimated in current budget projections.
What the Race Means for Technology Companies and Investors
The semiconductor sovereignty race creates distinct winners and losers across the technology ecosystem. Equipment companies — ASML, Applied Materials, Lam Research, KLA — benefit from every new fab construction programme regardless of geography, and their order books reflect the simultaneous fab buildout across the US, Europe, Japan, and India. Advanced packaging companies — Amkor, ASE, and the emerging US-based assembly operations funded by CHIPS Act provisions — are building capacity to serve the onshore manufacturing programmes that require domestic end-to-end supply chains for defence and government applications. Fabless chip designers — Qualcomm, AMD, MediaTek — face a more complex landscape where TSMC's geographic diversification creates both supply chain resilience and cost complexity that will eventually be passed through in chip pricing. The investor implication is that semiconductor equipment companies have the most durable exposure to the sovereignty capital cycle — every government's chip ambition requires their products regardless of which fabs ultimately succeed — while foundry investments carry geopolitical execution risk that is difficult to underwrite in standard corporate finance frameworks.