Europe is targeting gigawatt-scale Artificial Intelligence (AI) factories to bolster its lagging AI industry and address the challenges of a rapidly changing sector. Enthusiasm for the idea of factories implementing AI-driven production has gained momentum in recent months, especially after Nvidia CEO Jensen Huang emphasized the importance of infrastructure at a June event. Huang hailed a new “industrial revolution” at the GTC conference in Paris, France, stating that his company is working to help countries build AI factories that generate revenue through partnerships in France, Italy, and the UK. The ambitious goal of the European Union is to catch up with the U.S. and China, which currently hold significantly larger shares of global AI computing capacity.

The EU’s Strategy: Consolidating Resources and Billions in Investment

The European Union describes AI factories as a “dynamic ecosystem” bringing together computing power, data, and talent to create AI models and applications. The bloc has long lagged behind the U.S. and China in the race to scale AI. With 27 member states, the region is slower to act when it comes to agreeing on new legislation. Higher energy costs, permitting delays, and a power grid desperately in need of modernization are also factors that could hinder developments. Iliana Varkonnen, Deputy Vice President for Technological Sovereignty at the European Commission, noted that the bloc’s goal is to unify high-quality datasets, computing capacity, and researchers, all in one place. Varkonnen emphasized that the EU has all the necessary components to be competitive in this sector, while building technological sovereignty. She stated, “We have, for example, 30% more researchers per capita than in the U.S., focusing on AI. We also have about 7,000 startups developing AI, but the main obstacle for them is that they have very limited computing capacity. And that is why we decided, together with our member states, to invest in this vital infrastructure.”

To date, the EU has allocated €10 billion ($11.8 billion) in funding to establish 13 AI factories and €20 billion as a starting point for investment in gigafactories, marking what it defines as the “largest public investment in AI in the world.” The bloc has already received 76 expressions of interest for gigafactories from 16 member states across 60 sites, Varkonnen said. Varkonnen noted that the call for interest in gigafactories was “overwhelming,” far exceeding the bloc’s expectations. She stressed that these are “very large investments because they are four times more powerful in terms of computing capabilities than the largest AI factories, and that means billions in investment. Each of these requires an investment of three to five billion euros.” For the factories to constitute a significant addition to Europe’s computing capacity, a much more substantial investment from the private sector will be required to finance the expensive infrastructure.

AI Factories: From Raw Data to Advanced AI Products

The EU describes the facilities as a “one-stop shop” for AI companies. They are designed to mirror the process in industrial factories, which transform raw materials into goods and services. In an AI factory, raw data serves as input, and advanced AI products are the expected output. André Kuchnin, a stock research analyst at UBS, describes this as a “data center with additional infrastructure related to how the technology will be adopted.” Its purpose is to “create Graphics Processing Unit (GPU) capacity, which basically means building data centers with GPUs that can train models and run inferences… and then create infrastructure that makes it accessible to small and medium-sized businesses and parties who simply can’t build their own.” Martin Wilkie, a research analyst at Citi, adds that the manner in which the facility is used is key to its designation as an AI factory. According to him, “you create a platform by using chips with insane levels of computing capacity, and if you connect it to a network that can get the power to use them at full capacity, then the world is at your feet. You have tremendous ability to do something, but its success will be defined by what you use it for.”

Telecommunications company Telenor is already exploring potential use cases for such facilities with the launch of its AI factory in Norway last November. Telenor’s Chief Innovation Officer and Head of the AI Factory, Karen Hilsen, emphasized that data sovereignty is key. “If you want to use AI to innovate and make your business more efficient, then you might be putting critical, sensitive business information into these AI models,” she said. The company is working with BabelSpeak, which Hilsen described as a Norwegian version of ChatGPT, translating sensitive dialogues, such as in its pilot with border police, who cannot use public translation services due to security concerns. Hilsen stated that we are experiencing an “intelligence revolution” where “sovereign AI factories can really help advance society.”

Challenges and Prospects: Computing Power, Power Infrastructure, and Long-Term Outlook

Varkonnen said that the first AI factory in the region would become operational in the coming weeks, with one of the largest projects launching in Munich, Germany, in early September. But for gigafactories, it’s a different story. Bertin Martens, a senior researcher at Bruegel, questioned why such investments need to be subsidized by government funds. He noted, “We still don’t know how much private investment has been offered to complement the taxpayer subsidy, and what is the capacity and size of these factories. That is still unclear at this stage, so it is very difficult to say how much it will add in terms of computing capacity.”

Electricity consumption is also a key issue. Martens noted that building an AI-based gigafactory might take one to two years, but constructing a power generation plant of that size requires much more time. “If you want to build a state-of-the-art gigafactory with hundreds of thousands of Nvidia chips, you need to rely on at least one gigawatt of power consumption for one of those factories. Whether there is enough room in Europe’s electricity grid across all these countries to create these factories remains to be seen… It will require significant investment in electricity generation capacity,” he told CNBC.

UBS expects current global installed data center capacity, standing at 85 gigawatts, to double due to soaring demand. Based on the EU’s €20 billion investment and the plan for each factory to operate 100,000 advanced processors, UBS estimates that each factory could be around 100-150 megawatts, with a total capacity of approximately 1.5-2 gigawatts for all facilities. This could add about 15% to Europe’s total capacity—a considerable boost, even compared to the U.S., which currently holds about a third of global capacity. Following the announcement of the EU-U.S. trade framework, EU Secretary Ursula von der Leyen said on Sunday that U.S. AI chips will help power the EU’s AI gigafactories, aiming to help countries “maintain their technological edge.” However, Martens emphasized that “you could argue that it’s relatively easy, provided you have the money. It’s relatively easy to buy the chips from Nvidia and create these hardware factories, but to make it work and make it economically viable is a completely different question.” He said that the EU will probably have to start smaller-scale, as the region is not immediately capable of building its own AI models due to their cost. “I think over time, Europe will gradually be able to build its infrastructure and its business models around AI to reach that stage, but it won’t happen immediately,” Martens stated.

Conclusion and Forward Outlook

The ambitious efforts of the European Union to establish gigawatt-scale AI factories reflect a recognition of the critical need to bridge the gap with global AI powers. The significant public investments, coupled with a call for private sector involvement, indicate a deep commitment to building technological sovereignty. However, the challenges facing the EU are not simple, including the need for substantial private funding, resolving energy infrastructure issues, and managing the technological gap. Europe’s success in this project will depend on its ability to leverage its advantages (such as a large number of researchers) into a genuine competitive edge, and will be a significant indicator for the future of the global AI industry as a whole