World’s leading AI chipmaker makes infrastructure push across Europe
Nvidia, the world’s leading AI chipmaker, has unveiled a series of expansive partnerships across Europe, reinforcing its ambition to become the backbone of artificial intelligence infrastructure worldwide.
At its GPU Technology Conference in Paris on Wednesday, Nvidia CEO Jensen Huang announced a multi-country strategy involving governments, cloud service providers, telecommunications companies, and research institutions. The move aims to embed Nvidia technology deeply into Europe’s AI ecosystem, as the continent ramps up efforts to localize digital infrastructure and assert control over “sovereign AI.”
“Every industrial revolution begins with infrastructure. AI is the essential infrastructure of our time,” Huang said in a company release.
Nvidia’s European push includes infrastructure deals in France, Germany, Italy, Armenia, and beyond. A standout partnership has been inked with French AI startup Mistral, which will build a new AI cloud platform powered by 18,000 Nvidia Grace Blackwell chips. The platform is designed to accelerate AI adoption by giving businesses access to Mistral’s models and high-performance computing resources.
In Germany, Nvidia is rolling out what it calls an “industrial cloud,” a project centered around 10,000 Graphic Processing Units (GPUs) tailored to serve the needs of European manufacturers. Italy and Armenia are also set to host AI infrastructure projects, though further details are yet to be disclosed.
Telecommunications giants Orange and Telefónica are collaborating with Nvidia on AI applications, large language model deployments, and national-level digital transformation initiatives.
A central theme in Nvidia’s strategy is the concept of sovereign AI—ensuring that the data and servers powering AI applications remain within national or regional jurisdictions, especially within the EU. This emphasis responds to growing concerns over data privacy, national security, and digital dependency on foreign tech firms.
To support this, Nvidia is launching several new technology centers in countries including the U.K., France, Spain, and Germany, focusing on workforce upskilling, AI research, and scientific innovation.
Though Nvidia is globally recognized for its powerful GPUs, the company is also intensifying its software focus. It has expanded its DGX Cloud Lepton service—essentially a global GPU marketplace—by partnering with more cloud providers and integrating it with Hugging Face, a popular platform for AI model sharing.
In addition, Nvidia is promoting NIM (Nvidia Inference Microservices), a suite of pre-packaged AI models that developers can easily deploy. As of this week, any large language model available on Hugging Face can now be accessed via NIM, allowing developers to skip costly training phases and build applications more quickly. By tightly integrating its software offerings with its GPU infrastructure, Nvidia is positioning itself as a one-stop solution for AI deployment—a strategic move to outpace competition as global demand for AI accelerates.
