IA en un minutoNewsBusiness and products
NVIDIA reportedly plans to invest a further $2 billion in AI in Europe
IA en un minuto newsroom · Editor: Jon Elgezabal
In 30 seconds
NVIDIA wants to present within days new artificial intelligence investment deals in Europe adding up to $2 billion, Jensen Huang said in Paris. Huang also promised money for four French firms, among them Doctolib, where NVIDIA will come in as a minority partner. The tour moves on to London and Berlin, with open models offered as the recipe for Europe to gain ground.
NVIDIA expects to announce within the next week AI investment deals in Europe worth $2 billion. Jensen Huang, its chief executive, said so on Tuesday, October 13, at an event in Paris, according to Bloomberg as reported by Quartz. That figure comes on top of the at least $8 billion which, as Huang told a press conference covered by the AFP news agency, NVIDIA has put into Europe over the last year.
At the same event Huang also announced funding for four French startups: Doctolib, Entalpic, Genesis AI and Pigment. Doctolib, founded in 2013, supplies healthcare professionals with administrative tools and is a widely used service for booking medical appointments; Entalpic applies AI to chemistry and to research into materials; Genesis AI works on AI for robots, and Pigment makes business management software built on AI. In Doctolib's case NVIDIA will take a minority stake and the amount has not been made public, according to AFP.
Doctolib's chief executive, Stanislas Niox-Chateau, told AFP the money will let the company hire faster and speed up work on AI models of its own, meant for personal assistants serving patients and doctors and for lightening paperwork. He added that the company will run on NVIDIA chips the models it intends to keep control of, along with patient data, rather than relying on closed models such as those from OpenAI or Anthropic.
Paris opens a tour that also takes Huang to London and Berlin. NVIDIA holds its GTC conference in the German capital from October 20 to 22, according to The Next Web, and Huang is due to give a keynote there on Wednesday, October 21, at 11:00 Central European time, according to the NVIDIA blog.
At the event Huang presented open models (those whose parameters are published so others can use and modify them) as the way for Europe to close the gap with the main AI developers in the United States. "Open models are important as they can be customized for languages and for your own sovereignty," he said, according to Bloomberg. "Europe needs to get in the race," he added, in a remark carried by The Next Web.
Why it matters · analysis and opinion
Whoever sells the chips is also funding those who buy them. Doctolib, one of the chosen companies, says it will run its own models on NVIDIA chips, so the investment secures a customer at the same time. The case made for open models fits that logic: anyone who adapts an open model still needs machines to run it on. For Europe, the sovereignty message has a useful side and an awkward one. The French company wants models of its own so as not to depend on closed providers and to keep control of its patients' data, but that control ends where the chip maker begins. A business choosing a model is left with a practical question: which part of its system it really controls and which part stays in a supplier's hands. It remains to be seen which projects the promised money will go to.
Source: Quartz · Written with the help of AI: how we make the news
