AMD to acquire Fei-Fei Li's World Labs in $8.2 billion deal
Chipmaker AMD has agreed to acquire AI research company World Labs for $8.2 billion, with founder Fei-Fei Li joining AMD's leadership as the deal strengthens its position against rival Nvidia in AI chips.
AMD and World Labs announced an agreement under which the chipmaker will acquire the AI company for $8.2 billion. World Labs focuses on deep learning models designed to understand physical reality.
In a joint statement, World Labs said AI development requires close collaboration across model research, systems and compute. AMD said that understanding the kind of frontier workloads produced by World Labs will inform its chip development roadmap.
As part of the deal, World Labs founder Fei-Fei Li will join AMD as executive vice president and chief scientist. Li, a Stanford computer science professor, is widely regarded as a pioneer in AI and computer vision, known for her work developing the ImageNet database and its associated challenges.
AMD and World Labs had already formed a partnership last year focused on inference optimization and training, and the companies' ties have remained close since — Li appeared as a guest at AMD's CES presentation earlier this year.
Li founded World Labs in 2024 to build deep learning models with a stronger grounding in physical reality, arguing that genuine general intelligence requires understanding grounded in physics rather than text alone. In a post announcing the acquisition, Li said the company now has tangible proof of what's possible and wants to scale its efforts, widen its reach, and move closer to the hardware.
World Labs' first product, Marble, is aimed at creating entertainment experiences as well as simulated environments for training robots. The acquisition is expected to strengthen AMD's position against longtime rival Nvidia, which already offers a suite of open-weight world models called Cosmos, while AMD has so far only released text- and video-based models publicly.
World models are considered essential for training general-purpose robots, including autonomous vehicles and humanoid robots, given the shortage of real-world training data. The acquisition is expected to close before the end of the year, pending regulatory approval.
