IEEE Spectrum breed · Technologie
E.U.'s AI Ambitions Clash With Chip Dependency
The European Union's pursuit of technological sovereignty is complicated by its increasing demand for advanced semiconductors, essential for AI, while Europe's own chip production remains minimal and reliant on foreign designs and manufacturing.

The European Commission's Chips Act 2.0 aims to address this by introducing demand-side measures, such as public procurement, to stimulate domestic investment in chip design and manufacturing. This strategy, however, faces a paradox: the AI infrastructure Europe plans to build will initially depend heavily on processors designed by U.S. companies and produced in Asia.
The EU's AI Continent action plan includes numerous AI factories and gigafactories, alongside a significant expansion of data center capacity. These facilities will require a vast number of advanced AI processors, with estimates suggesting each AI factory needs up to 25,000 chips and a gigafactory at least 100,000. The majority of these are expected to be supplied by Nvidia, potentially creating an "Nvidia dependency trap" where Europe's computing infrastructure is technologically reliant on a single U.S. supplier.
Europe's reliance extends beyond Nvidia, encompassing the entire global chip supply chain. The U.S. dominates chip design and intellectual property, while advanced semiconductor manufacturing is concentrated in Asia, particularly Taiwan and South Korea. China plays a crucial role in materials and industrial processes. Europe's share in chip packaging, assembly, and testing is only 4 percent, with no top companies headquartered in the EU.
Despite these dependencies, Europe possesses strengths, including ASML, a leader in lithography systems, and Imec, a prominent semiconductor research center. However, these advantages do not guarantee autonomy across the entire value chain. Experts suggest that true sovereignty lies in resilience and reducing strategic vulnerabilities, not complete self-sufficiency, emphasizing the need for global collaboration, diversified suppliers, and strengthened domestic capabilities.
The success of Chips Act 2.0 will hinge on its ability to diversify suppliers and avoid shifting dependence from Asian manufacturers to U.S. technology companies, thereby mitigating geopolitical risks associated with reliance on a single country, company, or technology.
AI-samenvatting op basis van de bron.
IEEE Spectrum breed