Skip to main content
v2026.11,986 entries · CC-BY 4.0

Editorial · CASRAI · AI and ML research outputs

AMD completes World Labs deal and names Fei-Fei Li chief scientist

AMD announced the World Labs deal completion on November 16, bringing Fei-Fei Li and spatial AI researchers into its compute development.

Published 17 Nov 2026· 5 minute read

Ask CASRAI · free to try

Ask about this story

2 free questions a day, no account.

An AI assistant for research: funder rules, compliance, methods and publishing. It links the sources behind its answers and labels answers that come from a web search.

Answers draw on CASRAI's guides and dictionary plus the federal and funder documents we index: Federal Register, Grants.gov, Regulations.gov and UKRI.

Works on this site and inside Claude, Cursor and the AI tools you already use.

Everything CASRAI publishes — this page, the dictionary, the guides and the news — stays open to read, with no account and no card.

CASRAI is the reference for research administration — bookmark it for the next question.

AMD announced in Santa Clara on November 16, 2026 that it had completed its acquisition of World Labs, bringing Fei-Fei Li and the spatial AI research team into its computing business. The disclosure puts foundational model research alongside semiconductor engineering in AMD’s development plans. A filing published the same day records the legal closing as November 13.

Fei-Fei Li joins AMD’s scientific leadership

Li joins as executive vice president and chief scientist, reporting directly to AMD chair and chief executive Lisa Su. In its completion announcement, AMD says World Labs will continue foundational model research while working with its engineers on future hardware, software and systems. That remit connects the acquired team’s foundation model work to the infrastructure on which emerging AI applications run.

Su said the next stage of AI infrastructure requires companies to “optimize compute platforms and models together.” AMD describes the acquisition as supporting its open AI ecosystem strategy, with model research helping inform infrastructure for customers and developers. This is the company’s stated development direction.

The filing confirms the closing date and share payment

The November 16 Form 8-K/A says AMD issued 14,978,322 common shares to former World Labs stockholders at the November 13 closing. The merger agreement was signed on September 26, and AMD announced it on September 28. The previously disclosed purchase price was approximately $8.2 billion, payable in AMD shares and subject to adjustment.

The amended filing supplies the final share count, which had been undetermined in September. It says that count was calculated using AMD’s daily volume-weighted average share price over ten consecutive trading days ending on the second trading day before closing. The consideration goes to the acquired company’s former stockholders; the acquisition figure is not a newly announced cash allocation for research.

These dates distinguish the new disclosure from the earlier agreement: November 16 marks AMD’s public confirmation and amended filing, while November 13 is the transaction’s completion date.

Spatial AI brings different research workloads

AMD’s September announcement described San Francisco-based World Labs as developing models that generate, reconstruct and simulate interactive three-dimensional environments from text, images and video. It also identified robotic learning and simulation technology. AMD linked the acquisition to the expanding compute demands of reasoning, robotics, simulation and physical AI, explaining why model expertise mattered to its technology planning.

World Labs’ recent product activity supplies context for that research focus. Its Research and Insights index dates the Atlas tools and developer API release to November 12. The Atlas page describes Viewfinder for reconstructing scenes and exploring camera viewpoints, Chisel for guiding scene structure, and spatial environments for robotics research and simulation. CASRAI’s earlier report on the Atlas release covers those workflows. The acquisition disclosure concerns the organisation developing that technology and its role inside AMD.

In its July 28 research account, Building Worlds That Train Robots, World Labs describes a real-to-sim-to-real process: reconstruct a physical task, vary simulated conditions, and compare robot policies in simulation and on hardware. The account identifies object configurations, viewpoints, appearance and physical properties as controllable variations. It also describes evaluating where policies succeed or fail and whether simulation preserves their relative ranking. Those activities illustrate the training and evaluation work behind the acquisition’s reference to robotic simulation.

Company statements and independent assessments

When the agreement was announced in September, Li said joining AMD would give World Labs the resources and engineering depth to accelerate research and help define future infrastructure. Her statement placed collaboration among model researchers, systems specialists and compute engineers at the centre of the proposed combination. The completion announcement now confirms her AMD role.

MEKAWIRE Desk’s November 16 report interprets the purchase as acquiring a model team to inform the chip roadmap, with robotics representing a research workload. In a November 17 analysis, XenoSpectrum’s Y Kobayashi similarly stresses the application of research to hardware and software development, arguing that acquiring researchers alone does not establish a competitive improvement. These are assessments of the integration’s significance, rather than measured outcomes from the combined team.

What the integration means for research

Analysis: The concrete change for researchers following world models is where that research now sits. A team studying spatial environments and robotic learning has a formal place within a compute supplier’s scientific leadership, with an explicit mandate to collaborate with engineers. That creates an organisational route for findings about model workloads to influence infrastructure design. For university groups studying these workloads, subsequent research papers and documented implementations would provide evidence of how the collaboration affects their experiments.

AMD’s open-ecosystem language should also be read at the level at which it is stated: a strategy for infrastructure development. Assessing an open-source model requires the terms and materials of a particular release. The research implication is to evaluate model availability, implementation details and evidence of hardware support separately when outputs from the combined organisation appear.

Sources

Referenced across the research world

University of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logoUniversity of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logo
  • University of Cambridge logo
  • Columbia University logo
  • Crossref logo
  • University of Edinburgh logo
  • Harvard University logo
  • University of Oxford logo
  • Princeton University logo
  • Stanford School of Medicine logo
  • University College London logo
  • ORCID logo

View CASRAI adoption →