Europe sets its sights on multi-billion-euro gigawatt factories as it plays catch-up on AI

In a strategic move to enhance its artificial intelligence sector, Europe is focusing on the establishment of gigawatt factories. This initiative aims to address the continent's current lag behind the U.S. and China in AI advancements. The concept of AI manufacturing facilities has gained momentum, particularly after Nvidia CEO Jensen Huang emphasized their significance during a conference in Paris. He described this as a potential 'industrial revolution' and expressed Nvidia's commitment to partnering with countries like France, Italy, and the U.K. to develop these revenue-generating AI factories. The European Union envisions these factories as a 'dynamic ecosystem' that integrates computing power, data, and talent to produce innovative AI models and applications. Despite having a wealth of researchers and startups, the EU faces challenges such as high energy costs and regulatory delays that hinder its ability to scale up AI capabilities. Henna Virkkunen, the EU's executive vice president for technology sovereignty, highlighted the region's robust pool of researchers and startups but pointed out their struggles with limited computing resources. To combat this, the EU is investing significantly in essential infrastructure, pledging 10 billion euros for the creation of 13 AI factories and an initial 20 billion euros for gigafactory investments. The response to the EU's call for expressions of interest in these gigafactories has been overwhelming, with 76 submissions from 16 member states across 60 sites. However, achieving substantial computing capacity will require considerable private sector investment to complement public funding. The EU's vision for these facilities includes transforming raw data into advanced AI solutions, akin to traditional industrial manufacturing processes. According to Andre Kukhnin, an equity research analyst at UBS, the objective is to create GPU capacity within these data centers, enabling the training and inference of AI models while making these resources accessible to small and medium enterprises. The successful deployment of these facilities hinges on their usage and the ability to harness their immense computing power effectively. Telecommunications company Telenor is already paving the way in this domain, having launched its own AI factory in Norway. With an initial cluster of GPUs operational, Telenor aims to explore market opportunities before expanding further. Chief Innovation Officer Kaaren Hilsen emphasized the importance of data sovereignty in leveraging AI to drive innovation and efficiency in business. Virkkunen announced that the first AI factory is expected to commence operations soon, with a major project set to launch in Munich, Germany. However, the establishment of gigafactories poses substantial financial implications, requiring investments of three to five billion euros each. As the EU prepares to form a consortium for this initiative, questions remain about the level of private investment and the overall capacity of these facilities. Experts like Bertin Martens raise concerns regarding the necessity for government subsidies to fund these projects, as the specifics of private sector contributions remain unclear. Additionally, the power requirements for these ambitious projects present a significant challenge, with the construction of power generation facilities potentially taking longer than that of the gigafactories themselves. As the EU strives to enhance its data center capacities, projections indicate a potential increase of 15% to Europe's total computing capacity, contingent on effective execution and resource allocation. The collaboration with the U.S. on AI chip technology further underscores Europe's efforts to secure its technological future. Nevertheless, experts caution that while establishing hardware factories may be straightforward, ensuring their economic viability will be a complex endeavor. The consensus is that Europe needs to gradually build its infrastructure to achieve its ambitious AI goals over time.