AMD to Acquire World Labs for $8.2B, Betting on Spatial AI for Robotics

Advanced Micro Devices Inc. announced today that it will acquire World Labs Inc. for $8.2 billion through an all-stock transaction. The purchase follows AMD's participation in a $1 billion Series B funding round for the startup just seven months earlier, when the company was valued at $5 billion. Nvidia Corp. also participated in that earlier round and stands to gain substantially from the acquisition. AMD plans to finalize the deal before the end of the year.

World Labs was established in 2024 by Fei-Fei Li, a prominent artificial intelligence researcher renowned for developing ImageNet, a foundational dataset used to train object recognition algorithms. That dataset proved instrumental in the development of AlexNet, which laid the groundwork for contemporary AI systems.

Spatial Intelligence and Robot Training

The startup specializes in spatial intelligence models, commonly referred to as world models, which generate interactive three-dimensional simulated environments. These systems have particular relevance for robotics applications, where they enable the training of autonomous systems in virtual settings before real-world deployment.

Industrial robots typically rely on embedded AI models to execute their assigned tasks. During the training phase, developers place these models in simulated robotic scenarios and expose them to tasks mirroring those they will encounter in production. Through repeated attempts, the algorithms learn the most effective approaches to completing these assignments.

World Labs' technology accelerates the creation of these virtual training environments. The company's latest model, Atlas, can construct a detailed digital representation of a factory or warehouse using only a handful of photographs. The system also simulates dynamic elements like conveyor belts and furnishes depth information that helps AI systems better understand object locations and spatial relationships.

AMD's Robotics Strategy

AMD maintains substantial involvement in the robotics sector. The company invested $50 billion in 2022 to acquire Xilinx Inc., a leading manufacturer of field programmable gate arrays—specialized processors widely deployed in robotic systems. Recently, AMD introduced the Ryzen AI Embedded X100 series, a line of chips optimized for robotics applications. Each processor in this family integrates a central processing unit, graphics processing unit and AI accelerators, all housed in a hardened package engineered to endure extreme thermal conditions.

Broader Applications

Beyond robotics, Atlas offers capabilities in content creation and video manipulation. The model can synthesize videos and images from text descriptions and can integrate visual effects into existing video content.

Leadership and Future Direction

Building the compute platforms for the next generation of AI requires a deep understanding of how models are evolving. Fei-Fei and the World Labs team bring exceptional research leadership and model expertise. Together, we can use that insight to develop the hardware, software and systems that will power the next generation of AI and strengthen the open AI ecosystem.

AMD Chief Executive Lisa Su

Upon closing, Li will assume the position of executive vice president and chief scientist at AMD. The chipmaker has indicated that World Labs will maintain its focus on advancing AI research initiatives.