Nvidia and Japanese industrial coalition to build 140MW AI factory
Nvidia CEO Jensen Huang has secured agreements with a coalition of Japanese industrial firms to construct a 140MW 'AI factory' designed for physical AI development. The facility, slated for 2028, will utilize Nvidia's next-generation Vera CPUs and Rubin GPUs to train foundation models capable of large-scale video processing and synthetic media applications for robotics.
Key Takeaways
- Facility will house 13,750 Vera CPUs and 27,500 Rubin GPUs, targeting trillion-parameter model training by 2028
- Noetra consortium includes 44 domestic firms, led by SoftBank, Sony, NEC, and Honda, backed by $6.2B in government funding
- Robotics leaders including Fanuc and Yaskawa will adopt Nvidia's Cosmos 3 Edge models for on-device Jetson Thor processing
- Toyota expanded its partnership to include factory digital twins and urban-intelligence models for the experimental Woven City project
Why It Matters
The deals signal a shift from general-purpose generative AI toward 'physical AI,' where low-latency video processing and synthetic training data are applied to heavy industry. For the streaming and infrastructure sectors, this creates a massive new demand for high-bandwidth, edge-compatible video foundation models capable of real-world spatial awareness. It also marks Japan's attempt to establish a 'sovereign AI' stack, reducing reliance on U.S. or Chinese cloud dominance by housing compute and proprietary manufacturing data locally. Watch for whether this industrial-first approach accelerates the commercialization of multimodal models over pure-text LLMs.
Additional Context
The formation of the Noetra consortium follows a period of intensifying competition in the global robotics and foundation model market. Per Nikkei Asia in July 2026, Japan's Ministry of Economy, Trade and Industry (METI) committed approximately $6.2 billion to Noetra to ensure the country maintains its manufacturing lead as AI integrates into physical systems. This state-led strategy aims to deploy 10 million AI-equipped robots across 18 industrial sectors by 2040 to address critical labor shortages caused by a shrinking domestic workforce. Simultaneously, Nvidia's expansion into the Japanese automotive sector builds on foundations laid at CES in January 2025. According to Nvidia and Toyota, the world's largest automaker initially committed its next-generation fleet to the Drive AGX Orin platform. The July 2026 expansion significantly broadens this, incorporating Nvidia's Omniverse and Isaac Sim tools to create digital twins of assembly lines. Per Bloomberg in June 2026, these simulations allow manufacturers to train robotics in virtual environments before physical deployment, reducing operational downtime. Market analysts note that Japan is following a 'sovereign AI' blueprint similar to those recently adopted in Europe and Canada. Per Reuters in May 2026, several nations have begun subsidizing domestic compute clusters to protect industrial data from exposure to foreign-owned cloud platforms. For Japan, the Noetra factory represents the first national-scale infrastructure specifically optimized for the 'Vera Rubin' architecture, which offers tighter integration between CPU and GPU tasks for complex spatial reasoning.
Read full article at techcrunch.com
Get this in your inbox → Subscribe
Enjoy our coverage?
Add StreamingMeme as a preferred source on Google to see more of our streaming news at the top of your Search results.
Add as preferred source