Taiwan’s Foxconn has signed a new deal with Nvidia to create a new type of AI-focused data centre it dubs “AI factories”. The manufacturer says it will use the facilities to create a range of intelligence applications including self-driving cars.
Demand for AI applications has seen Nvidia become one of the most valuable companies in the world, as developers use its GPUs to train foundation models. As more models are trained there is also a growing need to find applications to capitalise on this new technology.
Foxconn believes the appetite for automation, machine learning and AI tools will lead to a new type of manufacturing. Nvidia CEO Jensen Huang calls this “the production of intelligence”, and says this will happen in “AI factories” that will operate at scale.
Announced during Foxconn’s annual tech showcase in Taipei, the collaboration will see a series of specialist data centres built using Nvidia chips and software and operated by Foxconn around the world. Huang gave the example of being able to process vast amounts of data from autonomous electric vehicles in real time to make them smarter.
“This car would of course go through life experience and collect more data,” said Huang while on stage with Foxconn chairman Liu Young-way. Holding a hand-drawn sketch of the potential factory, Huang said: “The data would go to the AI factory. The AI factory would improve the software and update the entire AI fleet. In the future, every company, every industry, will have AI factories.”
AI factories will use Nvidia ‘superchips’
The new AI factories will be built around Nvidia’s latest chips, including the GH200 superchip that is designed for large memory AI supercomputers and has been built to handle next-generation terabyte-class AI models for generative AI and graph analytics.
This includes the type of frontier AI models being discussed at the UK’s AI Safety Summit next month. These are thought to pose the greatest risk to security and society, due to the potential reach and scale. Models like Gemini from Google, OpenAI’s GPT-5 and Claude 3 from Amazon-backed Anthropic are likely to be made available in the coming months.
Foxconn is the largest supplier of Apple’s iPhones and a major technology hardware manufacturer. It is hoping to replicate the success and scale of its physical device operations with the new data centre and software project. This will be run alongside its growing electric vehicle business and demand for data analytics at scale.
Content from our partners
Reynolds Catering spurs innovation by upgrading its ERP to level up capabilities at scale
Zeelandia leverages AI to optimise precision, efficiency and pricing
How financial services companies can prepare for the EU’s DORA regulations
The company already has a partnership with Nvidia on autonomous vehicle platforms where Foxconn builds electronic control units for cars that are based on Nvidia’s own DRIVE Orin chip. Liu said Foxconn was “trying to convert itself from a manufacturing service company to a platform solution company”.
View all newsletters Sign up to our newsletters Data, insights and analysis delivered to you By The Tech Monitor team
The EV business seems to be driving much of this change at Foxconn. The company recently unveiled its Model N electric cargo van and confirmed it was in discussion with 14 potential customers to sell the vehicle in India and Japan.
Read more: AI adoption ‘surging’ across global defence sector
Topics in this article : AI , Foxconn , NVIDIA
