Technology Reported · Analysis

AMD is acquiring Fei-Fei Li’s World Labs for approximately $8.2 billion in stock

The deal brings a celebrated AI researcher and her 3D world-model technology to the chipmaker as it scrambles to close the gap with Nvidia.

Fei-Fei Li speaking at AI for Good 2017
Fei-Fei Li speaking at the AI for Good conference. File photo, 2017. ITU Pictures · CC BY 2.0 · via commons.wikimedia.org

AMD said Monday it is acquiring World Labs, the AI research lab co-founded by Dr. Fei-Fei Li, in an all-stock deal worth approximately $8.2 billion.

Li will join AMD as executive vice president and chief scientist, reporting to CEO Lisa Su. The acquisition is AMD’s second-largest on record, following the roughly $50 billion it paid for Xilinx in 2022, and it lands as the chipmaker chases Nvidia in the market for AI processors.

What World Labs brings

World Labs was founded in 2024 and reached a $1 billion valuation within months of launching. The lab is developing a so-called world model that can simulate 3D environments, a frontier AI category that goes beyond the text- and image-based systems that dominate today’s market. Its first commercial product, Marble, launched in 2025 and lets users create interactive 3D worlds from a prompt. Li and Su demoed Marble together at AMD’s CES presentation earlier this year, showing a model generating a 3D scene from a few images. “Intelligent agents, whether it’s robots or vehicles or even tools, can learn inside very rich physics-aware digital worlds before they even need to be deployed into the real one, making them much safer,” Li said at the presentation.

Li is a Stanford professor and a pioneer in computer vision, best known for building the ImageNet database and the AI competitions it inspired. She previously worked at Google, where she led AI research. Li has argued that true general intelligence requires a grounding in physics and the ability to understand and reason about data beyond text. After the deal closes, the World Labs team will continue to focus on advancing AI model research.

Closing the gap with Nvidia

AMD and World Labs formed an inference optimization-and-training partnership last year, and the acquisition deepens a relationship that already gave AMD insight into emerging model workloads. The chipmaker has ground to make up: Nvidia already has a suite of open-weight world models like Cosmos, while AMD has only offered text- and video-based models to the public. The stakes extend well beyond chips. Synthetic data from world models will be key for companies like Tesla and Figure to train general-purpose robots, because real-world training data for those systems is scarce and difficult to gather at scale.

The deal also arrives amid a wave of multi-billion-dollar AI acquisitions. Nvidia spent a combined $33 billion on Groq assets in December and Hugging Face this month, according to one source’s accounting of the two transactions. Meta spent $14 billion in 2025 for a minority stake in Scale AI, bringing in founder Alexandr Wang to lead a new AI division. OpenAI purchased Jony Ive’s startup io for about $6.4 billion last year.

The acquisition is expected to close by the end of the year, subject to regulatory approval. AMD plans to keep World Labs separate from its chipmaking business until the transaction closes. World Labs is based in San Francisco.

Corrections
No corrections have been issued for this article. Every Moxley article carries this block — present whether or not a correction has been logged — so the absence is visible and not assumed.

Sources & methods

  1. CNBC, September 28, 2026 Archived Sep 28, 2026
  2. The Verge, September 28, 2026 Archived Sep 28, 2026
  3. TechCrunch, September 28, 2026 Archived Sep 28, 2026

This article draws on reporting from three sources covering AMD’s announcement of its acquisition of World Labs. Figures attributed to a single source are noted as such; the disputed Hugging Face valuation is presented using the combined $33 billion figure from Source 1, per the editor’s instruction.

More from Technology