Is Europe Too Late to Compete in the Chips War?

October 21, 2025 Cristina Caffarra NO: Opening Statement at the Bloomberg Tech London 2025 debate, 20 October 2025 The topic of today’s debate is alas a common tech trope – especially from Americans and tech giants: “Europe is too late for X [replace X for any digital domain: hardware software connectivity], you hapless little Europeans let us do it all for you”. In chips the story typically goes: “Today Europe accounts for less than 10% of global semiconductors, it once produced nearly half of the world’s, it has fallen from grace and just cant catch up.” Indeed the loss of first mover advantage is the story of the European tech industry: we were pioneers, trailblazers, we are now a digital colony. The global spend in semiconductors in the next decade is estimated at $2-3 trillion, whatever the number turns out to be – Taiwan has the lion’s share incl $100bn in the US; the US itself accounts for over ¼ ($646 bn capex), growing share of fabrication from 10% to 14-15%, Korea investing massively also in memory (HBM), Japan in chips and wafers, and China racing for self sufficiency with giant state incentives to offset export controls, achieving major growth in legacy/mature chips. But two things: We are in a very different imperative today – with very changed perception of the sovereignty and security risk, just at the time when demand for advanced chips is expected to grow strongly because of (a) cars – as we shift to clean mobility (EVs), software-enabled cars, assisted/autonomous driving cars, (b) AI (c) defence. We have multiple design heroes supporting an ecosystem of flexible fabless companies, need I mention ARM in the UK, and multiple Research and Technology Organizations that go from early stage research to pre-industrialisation – French/Italian CEA-LETI in Grenoble, Belgian Imec, German Fraunhofer, Finnish VTT… Enters the king: Europe’s ASML has a global monopoly in advanced photolithography machines of incredible complexity, on which the whole world depends to produce the smallest chips at scale. Spend will more than double to 2032 and will at least preserve Europe’s 7-8-9% share of global foundry capacity – not just in the mid range but including plans for advanced nodes: Most likely scenario: expansion of the facility in Dresden below 5nm, through consortium involving Globalfoundries, Bosch, Infineon, NXP and TSMC, a planned investment of €10-15bn, could start production of 3nm nodes in 2027 Also discussions to create a European Boutique Fab for