EU revises Chips Act to focus on realistic semiconductor goals

Europe can compete in semiconductors, but not at every layer of the value chain and not by treating market-share targets as an industrial strategy. The more credible goal for Chips Act 2.0 is to make Europe harder to bypass in areas where it already has leverage: semiconductor equipment, power electronics, automotive and industrial chips, research infrastructure, photonics and selected advanced-packaging technologies. Catching Taiwan, South Korea and the US across leading-edge logic, memory, design and high-volume AI compute is a much less realistic proposition. This is the third part of our series on the proposed legislation. As eeNews Europe reported in Part 1, the European Commission tabled the new regulation on 3 June 2026. It remains a proposal under the ordinary legislative procedure, with work continuing in the Council and Parliament. Chips Act 2.0 meets Europe’s hard limits The starting point is uncomfortable. The European Court of Auditors concluded in 2025 that the EU was very unlikely to reach its original target of 20% of global semiconductor production by 2030. Its assessment put the likely share at about 11.7% and cited estimates that reaching 20% would require roughly €251 billion of expenditure. That is a useful reality check because the first Chips Act was never backed by an EU-level budget remotely comparable with the capital spending of the largest global chipmakers. The underlying industrial picture is uneven rather than uniformly weak. A Joint Research Centre analysis found that the EU is a strong net exporter of semiconductor manufacturing equipment, led by specialised wafer-fabrication machinery and lithography. It also put Europe at 10.6% of worldwide semiconductor shipments in 2023, with automotive demand accounting for a large share of the regional market. The same analysis points to import dependence in final chips, especially logic and memory. That distinction matters. Europe does not need to reproduce the entire global semiconductor supply chain inside its borders; no major economy does. But it does need enough capability at selected control points to avoid being merely a customer when supply chains tighten or geopolitical restrictions bite. Where Europe can still win Recent projects illustrate the difference between useful industrial capacity and headline-grabbing sovereignty claims. The European Semiconductor Manufacturing Company (ESMC) fab in Dresden is a genuine manufacturing gain. The TSMC-led joint venture with Bosch, Infineon and NXP is designed for 40,000 300 mm wafers per month using 28/22 nm planar CMOS and 16/12 nm FinFET processes. Production is targeted for late 2027 and full capacity for 2029. Those are valuable nodes for automotive and industrial applications, but they are not the frontier processes used for the most advanced AI accelerators. Infineon’s position is more revealing still. Its €5 billion Smart Power Fab in Dresden opened on 2 July 2026 and focuses on power semiconductors and analogue/mixed-signal technologies for automotive, industrial, energy and data-centre applications. That is precisely the kind of project Europe is structurally better placed to support: it expands an existing industrial strength, serves nearby customers and does not depend on pretending that every strategically important chip must be a leading-edge CPU or GPU. Advanced packaging is another plausible control point. The Commission approved €1.3 billion in Italian State aid for Silicon Box to establish an advanced packaging facility in Novara. Packaging cannot compensate for a shortage of front-end wafer capacity, but chiplets and heterogeneous integration make it increasingly important to system performance and supply-chain resilience. Research remains a European advantage, although it should not be confused with production. Imec’s NanoIC pilot line in Leuven gives industry access to beyond-2 nm process development, advanced interconnects and next-generation design kits. It is a formidable R&D asset, but a pilot line is not a high-volume foundry. Europe still has to convert research leadership into commercial manufacturing at scale. The cancelled Intel Magdeburg project shows the opposite risk. Intel abandoned the planned German megafab in 2025 after initially positioning it as a cornerstone of Europe’s leading-edge manufacturing push. As eeNews Europe reported when Intel cancelled the project, corporate strategy and demand can overturn years of subsidy planning. Semiconductor policy built around a few heroic mega-announcements is therefore fragile by design. Demand is the harder problem The Commission has recognised that subsidies alone are not enough. Its revised framework proposes permitting decisions within a maximum of 12 months for qualifying projects, new strategic-project designations and “Demand Accelerators” intended to connect chipmakers with European user industries. That is arguably the most important shift in the proposal. A fab is economically useful only if customers keep its tools busy. Europe has strong automotive, industrial, energy and communications customers, but it does not have the same concentration of hyperscalers, leading-edge fabless chip designers and AI accelerator demand that helps support frontier-node investment in the US and East Asia. Public money can narrow a cost gap; it cannot indefinitely manufacture a customer base. Three competitiveness scenarios Optimistic: Chips Act 2.0 passes broadly intact, the next EU budget provides a meaningful semiconductor funding line, permitting becomes materially faster and strategic projects are chosen around coherent regional clusters rather than national trophy projects. Europe becomes more indispensable in equipment, power electronics, photonics, advanced packaging and energy-efficient computing, while narrowing — but not eliminating — its leading-edge fabrication gap. Base case: Europe improves resilience in mature and specialty nodes, strengthens existing clusters and expands packaging and pilot-line capability, but its overall global share rises only modestly. The result is a stronger semiconductor economy without anything resembling full technological sovereignty. Pessimistic: state aid remains fragmented, operating costs stay uncompetitive, demand-side coordination produces little real purchasing power and multinational manufacturers continue to move projects according to global cycles. Europe ends up with excellent research infrastructure and several protected fabs, but little change in its bargaining power across the wider supply chain. Compete selectively, not everywhere The realistic case for Chips Act 2.0 is therefore narrower than the political rhetoric surrounding the first Chips Act. Europe can compete where it already has deep engineering capability, industrial customers, specialised infrastructure and defensible intellectual property. It can also use packaging, chiplets and photonics to gain leverage as system integration becomes more important. What it cannot do cheaply is recreate Taiwan’s foundry ecosystem, South Korea’s memory scale and the US design-and-hyperscaler complex at the same time. The ECA’s numbers make that plain. Success should be measured less by an arbitrary percentage of global production and more by whether Europe controls enough strategically important technologies and manufacturing capacity to remain a necessary participant in the semiconductor economy. The post Chips Act 2.0, Part 3: Can Europe compete? appeared first on eeNews Europe.