Let us know about free updates
Nvidia chief Jensen Huang said Europe’s shortage in the computing power of artificial intelligence will soon be “solved.”
Huang said at least 20 AI data centres are planned across Europe over the next few years, at the Vivedch conference held in Paris on Wednesday, predicting that the capacity of local data centres would increase by 10 times within the next two years.
This includes five “GigaFactories.” This includes sites with hundreds of thousands of Nvidia’s powerful graphics processing units. This is the chip required to train and run large language models.
“Europe is all moving towards AI,” fans said. “The amount of AI infrastructure built here will increase by several orders of magnitude over the next few years.”
Part of that growth comes from the expanded partnership between Nvidia and Mistral AI. This is a French startup that has emerged as Europe’s biggest hope for local developers, such as Openai and Deepseek, who can take on AI rivals in the US and China.
Mistral, an investor with Nvidia, will roll out the latest generation of the chipmaker’s Blackwell GPUs in its new facility. Mistral is building its first large data center in Essonne, just outside Paris, and plans to expand its infrastructure across Europe.
The announcement follows Huang’s warning on Monday that the UK is “missing” the infrastructure it needs to meet AI potential as Europe-based cloud providers NSCALE and NEBIUS announced plans to build new facilities using thousands of UK NVIDIA GPUs.
Transactions like these will bolster Nvidia’s efforts to promote “sovereign AI.” They are essential to diversifying their businesses from the small clutches of large American high-tech companies, which account for a huge share of their revenue.
Local capital and the availability of reliable power sources are seen as a barrier to European AI capabilities. Large AI facilities cost hundreds of billions of dollars to build and consume huge amounts of energy.
Recommended
A report from consultant firm McKinsey last October warned of “critical challenges” that will expand Europe’s AI capabilities, including limited power sources, lack of electrical engineering skills, and a slow planning process. McKinsey predicted that data centers in the region would need up to $300 million to invest in order to meet demand and almost triple power consumption by the end of the decade.
In Paris on Wednesday, Huang touted a massive expansion of “Indigenous AI infrastructure by European, European companies, European companies, European companies and European companies” over a data center being built by US cloud providers.
“Your AI shortage, your GPU shortage will be resolved for you soon,” he told researchers and startups in the Vivatech audience.
