AMD agreed on September 28, 2026, to acquire World Labs, the spatial-intelligence startup founded by Stanford computer science professor Fei-Fei Li, for $8.2 billion in an all-stock transaction expected to close by year-end1,2,3. Li will join AMD as executive vice president and chief scientist, reporting directly to CEO Lisa Su4. Co-founders Justin Johnson and Ben Mildenhall will continue to lead the World Labs team inside AMD14.
The deal is AMD's second-largest acquisition ever, after Xilinx. It comes seven months after AMD backed a $1 billion funding round for World Labs at a $5 billion valuation; Nvidia Corp. also participated in that round5,9. Bloomberg reported that Li's personal stake is now worth $1.6 billion10.
What World Labs builds
World Labs develops spatial-intelligence models that generate, reconstruct, and simulate interactive 3D environments for robotics and other physical-AI applications. Li founded the company in 2024 with Justin Johnson, Christoph Lassner, and Ben Mildenhall, raising $230 million in initial funding6. Li built the company on the thesis that true general intelligence required grounding in physics and the ability to reason about data beyond text.
The company's first publicly usable tool, Marble, generates small 3D spaces using Gaussian splats that can be exported as assets for film production, game development, and other applications. World Labs has since released additional models and tools, including ATLAS.
The unit formed after the acquisition will focus on building an open AI architecture integrating foundation models, software platforms, and specialized processing hardware11.
Partnership turned acquisition
AMD and World Labs formed an inference optimization-and-training partnership in 2025, working on model training and inference optimization on AMD GPUs. Li was a guest at AMD's CES presentation earlier this year. In a post announcing the deal, Li wrote: "Now that we have tangible proof of the possibilities, we want to do everything we can to accelerate the future". World Labs justified the acquisition by saying AI development required close collaboration across model research, systems, and compute. AMD said understanding frontier workloads, like those created at World Labs, will shape its chip-making roadmap.
Competitive stakes
One use case driving investment in spatial AI is generating synthetic data for training robots, a field where Nvidia Corp. holds a strong lead. AMD has produced its own AI models, including Micro-World, an open-source world model, but has not produced a stack that competes with Nvidia Corp.'s offerings in this area.
ANALYSIS Acquiring a research lab whose models stress both training and inference hardware gives AMD a captive workload to co-optimize silicon against, a tighter feedback loop than the arm's-length partnership the two companies maintained previously.
The $8.2 billion acquisition price represents a roughly 64% premium over World Labs' $5 billion valuation from the February funding round. The transaction is contingent upon regulatory approvals and customary closing conditions.