Europe is making a historic bet on AI independence by building seven massive computing factories across the continent, backed by over €20 billion in expected private investment alongside public funding. The European High-Performance Computing Joint Undertaking (EuroHPC JU) has opened applications for consortia to build and operate these AI Gigafactories, marking a major shift in how the EU plans to compete in the global AI race.

The initiative represents far more than just another data center project. These facilities will combine cutting-edge AI processors, supercomputing power, cloud infrastructure, and high-speed networks into integrated environments designed to serve researchers, startups, and established businesses across Europe. The goal is straightforward: give European companies and institutions the computing muscle they need to develop and train advanced AI models without relying on infrastructure controlled by companies outside the EU.

What Exactly Are AI Gigafactories, and Why Does Europe Need Them?

AI Gigafactories are industrial-scale computing facilities designed to handle the enormous computational demands of modern artificial intelligence. Unlike traditional data centers, these facilities are purpose-built for AI workloads, combining specialized hardware, software stacks, and support services tailored to the unique needs of AI development.

Each facility will be equipped with a very large number of state-of-the-art AI processors and will operate as a complete AI computing environment. Think of them as specialized factories for AI, similar to how semiconductor manufacturing requires dedicated facilities with precise environmental controls and specialized equipment. The difference is that these factories produce computing capacity and AI services rather than physical chips.

Europe's motivation is clear: the continent currently lacks sufficient sovereign computing infrastructure to train frontier AI models. This dependency creates strategic vulnerabilities. By building these gigafactories domestically, the EU aims to strengthen its technological independence while creating a competitive advantage for European researchers and businesses.

How Will These Facilities Support European AI Development?

• Training and Fine-Tuning Services: The gigafactories will provide the computational power needed to train large AI models from scratch and customize existing models for specific applications, services currently dominated by non-European providers.
• Inference Infrastructure: Beyond training, these facilities will offer inference services, meaning they'll run trained AI models at scale for researchers, businesses, and public-sector organizations across Europe.
• Accessible Support for Startups and SMEs: A particular focus will be supporting small and medium-sized enterprises and startups, which typically lack the capital to build their own computing infrastructure but need access to powerful AI tools to compete.
• Secure Cloud Access: The facilities will provide secure, standardized cloud access with specialized AI support services, making it easier for organizations to adopt AI without managing complex infrastructure themselves.

How Will Europe Select and Fund These Megaprojects?

The selection process is competitive and rigorous. Consortia and special purpose vehicles bringing together companies, public entities, investors, and other partners can apply. These groups can establish gigafactories in a single EU member state or create cross-border facilities spanning multiple countries, offering flexibility in how projects are structured.

The submission deadline is November 12, 2026, with proposals undergoing competitive evaluation to select successful projects. Following selection in early 2027, the gigafactories are expected to begin operations within 18 months, meaning some facilities could be operational by late 2028.

Funding comes from a clever public-private partnership model. The EU and participating member states will act as anchor customers, jointly procuring compute access time from the selected facilities. This de-risks the massive capital investment required and helps unlock the expected €20 billion in private investment across the EU. Some public funding will come from the current budget framework, but the bulk will come from the next Multiannual Financial Framework covering 2028 to 2035.

Different gigafactories will receive different levels of EU and national co-funding, reflecting the size and scope of sovereign services they'll deploy. This tiered approach allows for projects of varying ambition while ensuring that public money is allocated efficiently.

Where Will These Facilities Be Located?

The initiative will support the establishment of up to seven AI Gigafactories, each located in at least seven different EU member states. However, some gigafactories may operate as cross-border collaborations, deploying infrastructure across multiple countries. This geographic distribution ensures that computing capacity and AI services are available throughout Europe rather than concentrated in a single region.

This distributed approach also serves a strategic purpose: it strengthens the AI ecosystem across the entire continent, not just in wealthy or tech-forward regions. By spreading these facilities geographically, Europe can build a more resilient and inclusive AI infrastructure.

How Does This Fit Into Europe's Broader AI Strategy?

The AI Gigafactories initiative doesn't exist in isolation. It complements the AI Factories initiative, which is already overseeing the implementation of 19 AI factories across Europe, supplemented by 13 AI Factory Antennas that provide free, customized support to startups and small businesses.

Together, these initiatives represent Europe's comprehensive approach to building sovereign AI capability. The EuroHPC JU has already deployed 12 world-class supercomputers across Europe, including JUPITER in Germany and Alice Recoque in France, which are Europe's first exascale systems. The new gigafactories will build on this foundation, providing the next generation of AI-optimized computing infrastructure.

The EuroHPC JU's mandate was recently expanded under Council Regulation (EU) 2026/150 to include deployment of AI Gigafactories and advancement of quantum technologies. This expansion signals Europe's commitment to competing at the frontier of computing technology, not just in AI but across the entire spectrum of advanced computing.

What's at Stake for European AI Innovation?

The success of this initiative will determine whether European companies and researchers can compete globally in AI development. Currently, most frontier AI model training happens on infrastructure controlled by non-European companies, creating a dependency that limits European innovation and strategic autonomy. By building sovereign gigafactories, Europe aims to level the playing field.

The €20 billion in expected private investment signals confidence from the business community that these facilities will be economically viable and valuable. This isn't just public spending; it's a vote of confidence from private investors that there's a real market for European AI computing infrastructure.

For startups and SMEs, these facilities could be transformative. Currently, building and training AI models requires either massive capital or partnerships with large tech companies. Accessible, affordable compute infrastructure could democratize AI development across Europe, enabling a new wave of innovation from organizations that previously couldn't afford to compete.