Semiconductor giant AMD has officially completed its acquisition of spatial intelligence startup World Labs, bringing one of the industry's premier model research groups directly into its high-performance compute business . Following the transaction, World Labs co-founder Dr. Fei-Fei Li has joined AMD as Executive Vice President and Chief Scientist, reporting directly to Chair and CEO Dr. Lisa Su .

The closing marks a significant milestone in AMD's strategic expansion across the entire AI technology stack. By uniting World Labs' groundbreaking spatial models and 3D simulation tools with AMD's silicon and software engineering teams, the company aims to co-design hardware architectures and foundational models to power future superintelligence workloads .

AMD Finalizes $8.2 Billion World Labs Transaction

The deal was initially announced in late September as an all-stock transaction valued at approximately $8.2 billion . With the regulatory approvals and closing procedures now finalized, AMD has absorbed World Labs into its corporate structure, issuing approximately 14.98 million shares of common stock to former World Labs security holders as part of the transaction [AMD Agrees to Acquire Dr. Fei-Fei Li's World Labs to Expand AI Compute].

Under the terms of the agreement, the World Labs team will continue advancing its core research agenda in spatial intelligence, physical AI, and interactive 3D environments . Key research leaders Justin Johnson and Ben Mildenhall will continue guiding the specialized research group alongside Dr. Li, maintaining their focus on foundational 3D models while integrating closely with AMD's hardware and systems teams .

Dr. Fei-Fei Li Joins as Executive Vice President and Chief Scientist

Widely regarded as one of the godmothers of modern artificial intelligence, Dr. Fei-Fei Li brings decades of scientific leadership to AMD . Her appointment as Chief Scientist signals AMD's commitment to steering silicon design directly around the needs of next-generation generative models. Dr. Li previously co-directed Stanford University's Human-Centered AI Institute and founded ImageNet, the dataset that catalyzed the modern deep learning revolution.

Speaking on the completion of the transaction, AMD CEO Dr. Lisa Su highlighted the mutual benefits of aligning silicon design with AI research. Delivering the infrastructure that will power the next era of superintelligence requires optimizing compute platforms and models together, Su noted, adding that World Labs' pioneering model research will play a vital role in shaping future hardware, software, and systems across AMD's enterprise portfolio .

Co-Designing AI Models and Future Compute Infrastructure

The acquisition of World Labs reflects a growing shift in the semiconductor industry toward co-designing physical compute infrastructure alongside frontier AI models. Rather than developing graphics hardware in isolation, AMD plans to leverage real-world feedback from World Labs' complex workloads to optimize future Instinct GPU architectures, interconnects, and memory configurations .

World Labs specializes in spatial intelligence, creating AI models capable of understanding and generating interactive 3D spaces from 2D input . Models like Atlas allow systems to infer geometry, lighting, and physics in real time, serving as critical foundations for robotics simulation, spatial computing, and digital twin environments . These computationally intensive workloads require massive memory bandwidth and efficient inter-node communication, areas where semiconductor design decisions directly govern model training speeds.

By bringing this research unit in-house, AMD gains immediate visibility into the bottlenecks faced by frontier model developers. Industry analysts point out that tech hardware giants are increasingly acquiring software and model teams to ensure their silicon designs remain tailored for emerging AI paradigms, a trend also seen as competitive moves unfold across cloud data center providers like Microsoft Azure and chip competitors like NVIDIA.

Impact on Open Ecosystem Hardware and Software Development

A central tenet of the combined entity is a commitment to an open AI ecosystem . AMD has consistently positioned its ROCm open software stack as an alternative to proprietary developer ecosystems, working to lower barrier entry costs for researchers across cloud infrastructure and edge platforms.

Dr. Li reiterated this shared vision, noting that joining forces with AMD allows World Labs to scale its technological reach while getting closer to physical silicon . By publishing open model weights and optimizing spatial algorithms directly on AMD hardware, the partnership aims to provide developers with accessible tools for robotics, gaming engines, and industrial design simulation without locking researchers into single-vendor hardware environments .

Furthermore, the integration is expected to benefit consumer and developer hardware down the line. Beyond enterprise accelerators, hardware capabilities perfected through 3D simulation research are expected to inform consumer microarchitectures, embedded systems, and client-side processing platforms across AMD's broader product portfolio.

Looking Ahead

With the acquisition complete, AMD moves into its next phase of enterprise growth with an expanded leadership team and a strengthened software footprint . As physical AI and 3D spatial models become increasingly critical to commercial deployment, the synergy between Dr. Fei-Fei Li's research vision and AMD's silicon manufacturing ecosystem will be key to watching how compute architectures evolve in the years ahead.