Nvidia and Foxconn announced they are teaming up to create what they call ‘AI factories.’ These are special data centers that will be able to provide really powerful computing abilities. They aim to use these capabilities to speed up the development of self-driving cars, machines that can work on their own, and robots used in industries.
The collaboration between Nvidia’s founder and CEO, Jensen Huang, and Foxconn’s chairman and CEO, Young Liu, was announced at an event in Taiwan.
The AI factory will be based on Nvidia’s powerful computing technology, and it will be used to handle, refine, and convert large amounts of data into valuable AI models and information.
Nvidia and Foxconn’s Full AI System for Self-Driving Cars
Huang explained that they’re creating a complete system. On one side, they’re making advanced electric vehicles (EVs) with artificial intelligence (AI) that allows them to interact with drivers and passengers and also drive autonomously.
On the other side, there’s an AI factory responsible for developing software for these cars. The car gathers data as it goes through life experiences, which is then sent to the AI factory. At the factory, the software is improved and updated for the entire fleet of AI-driven cars.
The AI factory collaboration between Nvidia and Foxconn is an extension of their previous partnership announced in January. This earlier agreement involved Foxconn becoming a primary supplier of electronic control units (ECUs) for car manufacturers.
These ECUs are made using Nvidia’s Drive Orin system-on-a-chip (SoC), which is a powerful AI platform for autonomous driving. On Tuesday, Foxconn also committed to producing ECUs with Nvidia’s next-generation SoC, Drive Thor, starting in 2025.
As part of this partnership, Foxconn, known for offering ready-made electric vehicle (EV) platforms for car manufacturers, stated that the vehicles they produce as a contract manufacturer will incorporate Nvidia’s Drive Hyperion 9 platform.
This platform includes not only Drive Thor but also a set of sensors like cameras, radar, lidar, and ultrasonic sensors, which are essential for enabling self-driving capabilities in the vehicles.
Making Electric Cars and the Growth of AI Factories
Foxconn already has a contract to manufacture electric vehicles (EVs) for Fisker, despite facing a lawsuit from its former partner, Lordstown Motors. For Foxconn to make its AI factories successful, especially when competing with Tesla, it needs to operate on a large scale.
These AI factories are essentially competitors to Tesla’s Dojo supercomputer, which Elon Musk’s company began producing over the summer. Dojo is responsible for training Tesla’s neural networks, which are essential for powering, training, and improving their advanced driver assistance system known as “full self-driving” (FSD).
Elon Musk, hopes that FSD will eventually achieve complete self-driving capability, and this is where the incredible computing power of Dojo comes into play.
Currently, Tesla uses a large supercomputer based on Nvidia’s GPUs (Graphics Processing Units).
However, the new Dojo will be uniquely designed with chips created by Tesla. On the other hand, the Foxconn-Nvidia AI factories will be based on Nvidia’s GH200 Grace Hopper Superchip and AI Enterprise software, as stated by the company.
These factories will be used for more than just self-driving cars.
During the tech event at Hon Hai, Liu expressed Foxconn’s ambition to transform itself from a company that primarily offers manufacturing services into one that provides platform solutions. They plan to expand the AI factories into various industries, starting with three key platforms: Smart electric vehicles (EVs), smart cities, and smart manufacturing.
Huang explained, “This is a facility that takes data as input and generates intelligence as an outcome.” Liu agreed, and they believe that in the future, every industry and every company will have its own AI factory.
Nvidia and Foxconn are teaming up to build AI factories that promise to speed up progress in self-driving cars, autonomous machines, and industrial robots. They’ll use Nvidia’s powerful computing tech to process data and create valuable AI information. This collaboration expands their previous deal to supply electronic control units for automakers. These AI factories will impact many industries, not just automobiles, as they’re a move towards integrated AI solutions.