AI
Jensen Huang's Japan Tour Locks in Physical AI Deals Across Robotics and Chip Supply
Nvidia's CEO secured partnerships spanning Japan's industrial base, from a government-backed AI factory to alliances with the country's top robotics firms.
TechCrunch — AI
Jensen Huang spent July 15–16 in Tokyo on a strategic sweep through Japan's tech establishment, emerging with commitments that position Nvidia's chips at the core of the country's pivot toward factory automation and robotics. The visits—following earlier stops in Taiwan and South Korea—signal Nvidia's determination to shape the industrial AI frontier across Asia's manufacturing heartland.
The trio of initiatives Huang secured reveals Japan's multi-pronged bet on what Nvidia calls physical AI: software and systems designed to run machines, robots, and factory floors rather than just process information.
Japan's National AI Play: Noetra#
Tokyo is backing its own homegrown AI infrastructure to avoid dependence on American or Chinese models. The government assembled roughly 44 domestic firms—anchored by SoftBank, Sony, NEC, and Honda—under an initiative called Noetra, committing up to 1 trillion yen ($6.2 billion) over five years.
The hardware, however, is unmistakably American: Nvidia is building what it calls "a Vera Rubin AI factory"—a data center equipped with 13,750 Vera CPUs and 27,500 Rubin GPUs, projected to deliver 140 megawatts when it launches in 2028. The initiative follows a three-phase roadmap: a Japanese-language-focused reasoning model in fiscal 2026, a multimodal variant handling text, images, video, and audio by 2028, and specialized "Real-world Native AI" for robots by 2030.
Japan's Robotics Coalition Embraces Nvidia's Cosmos#
Nvidia introduced Cosmos 3 Edge, a lightweight variant of its physical AI foundation model, to a coalition of Japan's premier industrial names: Fanuc, Yaskawa, Kawasaki Heavy, Fujitsu, Hitachi, NEC, Sony, SoftBank, Kubota, and the robotics alliance AIRoA. Cosmos 3 Edge runs on Nvidia's Jetson Thor chips embedded directly inside machines, removing the need to route data back to centralized servers.
Huang framed the opportunity in sweeping terms: "The next frontier of AI is in the physical world, and this is a once-in-a-generation opportunity for Japan," he . Several partners are already piloting shared control systems; companies like Honda R&D and Omron are building production-ready tools.
