France is making a big bet to become Europe’s top AI hub. A consortium led by the cloud company Scaleway is pushing to build one of the European Union’s five planned AI gigafactories. Their bid is around $10 billion, focused on a huge data center powered by next-generation GPUs.

This facility would pack the computing power of more than 288,000 Nvidia H100 GPUs. That would make it the largest single-country AI cluster outside the US hyperscalers like Microsoft and Google. The plan centers on advanced GPU clusters designed to train and run cutting-edge AI models.

The consortium, called AION, includes a broad mix of French AI companies and research groups. Partners range from chip designers like SiPearl to AI software platforms like Hugging Face. It also links to national research centers such as Inria and GENCI. This team blends private industry with public research, aiming to build a European AI infrastructure that’s sovereign and open.

Behind the scenes, the European Commission launched a €20 billion program to fund up to five AI gigafactories across Europe. This is part of a bigger push to reduce dependence on US and Chinese tech giants. The goal is to secure Europe’s AI future with homegrown technology and infrastructure.

SoftBank’s $100 Billion Ambition in France

Alongside the French consortium, SoftBank is eyeing a massive AI data center investment in France. Talks between SoftBank’s CEO Masayoshi Son and French President Emmanuel Macron suggest a potential $100 billion project aimed at making France a strategic AI hub.

This project, if it happens, would vastly expand Europe’s AI computing power. It would also reinforce France’s role in the global AI race. SoftBank already invests heavily in AI infrastructure worldwide. This new plan could link French AI efforts to SoftBank’s global network.

The discussions are private, but an official announcement could come soon. Macron raised the idea during a visit to Tokyo. The potential project aligns perfectly with France’s goal to attract large-scale AI investments and build sovereign technology capabilities.

Europe’s Race for AI Data Centers and Sovereignty

Europe is racing to build AI data centers quickly. Governments offer fast permits, tax breaks, and cheap land to attract these facilities. Italy, Germany, and Spain have all changed laws to speed up approvals and encourage investment.

Meanwhile, in the US, big AI data centers face pushback. Communities protest over water use, energy costs, and land rights. Some projects have been delayed or canceled because of local opposition. Europe has avoided this backlash so far by smoothing the path for new construction.

However, Europe faces a critical challenge. Its AI computing depends heavily on GPUs made by companies outside Europe, mainly Nvidia and Taiwan’s TSMC. True chip independence remains out of reach. This reliance on foreign hardware limits Europe’s technological sovereignty.

The French-led AION project tries to address this by combining open-source AI models and European-designed chips. It uses hardware from French companies like SiPearl and Eviden, paired with software from local AI labs. This mix aims to create a more self-reliant AI ecosystem.

The European AI gigafactory program plans to build mega data centers that dwarf existing AI facilities. Each gigafactory might hold around 100,000 GPUs and serve startups, researchers, and companies across the continent. The goal is to boost Europe’s competitive edge and reduce reliance on US cloud providers.

Still, questions remain about who will actually use these huge computing hubs. Some experts worry they might favor big companies over smaller innovators. The European Commission and member states will need to balance access and control carefully.

For now, France is leading the charge. Its $10 billion AION bid and SoftBank’s potential $100 billion investment show strong political and industrial will. With energy from a low-carbon grid and a strong tech ecosystem, France hopes to become Europe’s AI powerhouse.