AMD is making a major push beyond AI chips, agreeing to acquire Fei-Fei Li’s World Labs for approximately $8.2 billion in an all-stock transaction, strengthening its ambitions in spatial intelligence and physical AI.
World Labs develops AI models capable of understanding, generating and simulating interactive 3D environments. Its technology is designed to help machines understand how objects, spaces and actions behave in the physical world.
Key Highlights
-
$8.2 billion acquisition: AMD will acquire World Labs through an all-stock transaction.
-
Fei-Fei Li joins AMD: She will become Executive Vice President and Chief Scientist.
-
Physical AI focus: World Labs brings spatial intelligence and 3D world-model expertise applicable to robotics, simulation and autonomous systems.
-
Beyond chips: AMD is strengthening its position as a full-stack AI company by connecting frontier model research directly with future computing infrastructure.
The acquisition brings together two influential AI leaders: AMD CEO Lisa Su and World Labs co-founder and CEO Fei-Fei Li, the computer-vision pioneer whose work helped accelerate modern deep learning.
Following completion of the transaction, Li will join AMD as Executive Vice President and Chief Scientist, reporting directly to Su and helping shape the semiconductor company's future AI technology strategy.
World Labs' technology could become particularly important as AI expands beyond chatbots into robotics, autonomous systems, simulation, industrial design and scientific applications.
Its latest world model, Atlas, can work across text, images, video and 3D information to generate, reconstruct and simulate environments, providing machines with richer spatial understanding.
AMD had already invested in World Labs and collaborated with the startup to optimise model training and inference using AMD GPUs, giving the companies an established technical relationship before the acquisition.
Strategically, the deal could help AMD design future GPUs, software and computing systems around emerging AI workloads rather than simply supplying processors after new models have been developed.
It also highlights the semiconductor industry's next competitive frontier. Nvidia and AMD are increasingly positioning themselves not merely as chip manufacturers but as full-stack AI infrastructure companies spanning hardware, software, models and ecosystems.
The acquisition is expected to close by the end of 2026, subject to regulatory approvals. More importantly, it signals where AI may be heading next: from understanding language to understanding the physical world—and eventually acting intelligently within it.
