Nvidia CEO Predicts 1000% AI Power Surge in Europe by 2025

During the VivaTech technology summit in Paris, Nvidia CEO Jensen Huang predicted that Europe's AI computing power will increase tenfold over the next two years. This projection is backed by the continent's plans to build over 20 "AI super factories," which are expected to have a power capacity exceeding 1 gigawatt for large-scale data centers. This significant investment in AI infrastructure is poised to accelerate the development and deployment of advanced AI technologies across Europe, potentially positioning the region as a global leader in AI innovation.

Huang emphasized that Europe has recognized the importance of these AI factories. The tech giant, which provides chips for most global AI computing projects, noted that multiple large-scale data centers under construction will have a power capacity exceeding 1 gigawatt, aiming to rank among the world's top data centers. Huang also highlighted that researchers and startups facing AI resource and GPU shortages will soon see these issues resolved, with a bright future ahead.

Despite lagging behind the U.S. in AI infrastructure and investment compared to other regions, Huang stressed during a meeting with British Prime Minister that inadequate infrastructure is hindering the UK's ability to leverage its talent and startup vitality in the AI field. This European tour is a crucial part of Huang's global strategy to promote AI technology applications.

While specific investment figures were not disclosed, Huang estimated that the construction cost of a single AI factory could reach 500 billion dollars. The European Union has announced a 20 billion euro special fund to build five AI super factories equipped with over 100,000 processors, focusing on the development and training of next-generation AI models.

During his time in Paris, Nvidia announced several European AI infrastructure cooperation plans, including a partnership with French AI startup Mistral to build an "AI cloud platform." This service, named Mistral Compute, will be powered by 18,000 of Nvidia's latest Grace Blackwell chips and will be initially based in the data center in Essonne, France, with plans to expand to other European regions. French President Macron praised this collaboration, calling it a key step in defending digital sovereignty and strategic autonomy, and urged European companies to actively adopt this AI computing service.

To expand its AI accelerator market, Nvidia is promoting the construction of national-level AI infrastructure in various countries. Huang emphasized to reporters that each country should establish an independent AI system, with data being the foundation of AI—the key is not to export data but to purchase equipment for independent development. Currently, companies like Nebius and Nscale in the UK, as well as Italy and Armenia, are accelerating hardware upgrades. Nvidia's European ecosystem has attracted 1.5 million developers, 9,600 enterprises, and 7,000 startups to participate in its "Startup Acceleration Program."

Nvidia's data center and high-performance computing director Dion Harris stated in a pre-speech briefing that Europe's only weakness in catching up with the AI wave is its infrastructure. He mentioned that Nvidia is deeply cooperating with European cloud service providers and telecom operators. On Wednesday, Nvidia also announced the creation of the world's first "industrial AI cloud," with a German base integrating tens of thousands of GPUs and multiple systems and servers to simulate the entire process from wind tunnel experiments to intelligent car design.

Over the past three years, this Silicon Valley giant's quarterly revenue has approached its rival Intel's annual performance, with its AI accelerator chips dominating the market. Customers include tech giants like Microsoft and Meta. Now, Nvidia is further expanding its market share by promoting the use of miniaturized systems by enterprises and nations. Through the Lepton cloud service, Nvidia is connecting AI developers with computing resources, with new partners like Amazon AWS and Mistral further expanding its service network. Previously, cloud computing companies like CoreWeave and SoftBank have used this service. The company is also focused on helping various European countries deploy AI models based on local language data.

In the field of intelligent driving, Mercedes-Benz CLA models equipped with Nvidia technology are already on the road, and new models from Volvo and Jaguar will also soon feature its Drive platform, marking the commercialization of years of technological accumulation.