According to a detailed report by TechCrunch, Nvidia CEO Jensen Huang conducted an intensive two-day working visit—on July 15 and 16, 2026—to Tokyo, Japan’s capital. During the visit, Huang met with representatives of the nation's industrial elite and the leaders of its semiconductor supply chain. This visit took place just weeks after he delivered a keynote address in Taiwan, and months after visiting South Korea, where he signed a major deal to supply 50,000 GPUs in Seoul last fall. Huang left Tokyo having secured agreements that span Japan's entire technological ecosystem—from establishing a national artificial intelligence factory and forging strategic partnerships with the country's leading robotics companies, to signing agreements with chip-material suppliers that power Nvidia's next-generation AI processors.
His message throughout the trip was unequivocal: Nvidia is targeting Japan’s factory floor, and many of the country’s largest manufacturers are joining this effort. According to Huang, the next and future chapter of AI belongs to factory floors, robots, and machines, and he wants Japan to be the nation that leads and builds it. The history between the company and the country is long-standing. Thirty years ago, a $5 million investment from the Japanese gaming company Sega helped keep Nvidia alive when it was on the verge of bankruptcy. Today, Nvidia and Japan’s industrial giants need each other once again—this time to build the physical-AI era, starting with three major, leading projects.
Noetra — Japan’s Sovereign AI Play
The Noetra project represents Japan's sovereign move in the field of artificial intelligence (sovereign-AI). According to the TechCrunch report, Japan does not want to run its factories and robots on AI systems developed in the United States or China. For this reason, the Japanese government mobilized and gathered approximately 44 leading domestic firms, with giants SoftBank, Sony, NEC, and Honda standing at the core of the venture. Their common goal is to build their own independent AI for robots, vehicles, and factory floors. The Tokyo government commits to allocating up to 1 trillion yen (equivalent to approximately $6.2 billion) over a five-year period for this purpose. This is a significant bet on domestic "physical AI," based on foundation models designed to power machines and devices. While Japan wants to own full ownership of the software brain, the hardware required to build this system still comes from the American company Nvidia.
The US chip giant is building the "Vera Rubin AI factory," a massive data center packed with the company's next-generation chips, which is expected to officially launch in 2028. This data center will house approximately 13,750 Vera CPUs and 27,500 Rubin GPUs, delivering an impressive 140 megawatts of power. The Noetra organization will oversee this effort and plans to physically build the data center.
Noetra's strategic plan is slated to be executed in three clear, successive phases:
- The first phase will focus on developing a reasoning model with a heavy emphasis on Japanese-language skills, expected to begin in fiscal year 2026.
- The second phase will involve developing an omni-modal version capable of handling text, images, video, and audio by 2028.
- The third phase will yield a system called "Real-world Native AI," built specifically to directly run robots by 2030, which will be released to developers outside the Noetra organization in defined phases.
Japan’s Robotics Coalition Around Cosmos Models
In parallel with the sovereign project, a broad robotics coalition is forming in Japan as the country's industrial giants align behind Nvidia's Cosmos platform. According to the report, Nvidia is aiming directly at Japan's factory floors, and many of the country's leading robotics and manufacturing players are signing cooperation agreements. Giants Fanuc, Yaskawa, Kawasaki Heavy Industries, Fujitsu, Hitachi, NEC, Sony, SoftBank, Kubota, as well as the robotics group AIRoA, have all declared their intention to build their developments on top of Nvidia's Cosmos models. This is an open model that Nvidia launched in May in partnership with a handful of leading AI labs worldwide.
During his official visit to Tokyo, Nvidia gave these companies a compelling reason to commit to the project when it unveiled Cosmos 3 Edge. This version of the model is designed to run directly on Nvidia’s Jetson Thor chips, which are installed inside the machines and robots themselves on-site. Some of the participating companies have already begun testing a shared control and management system on the ground; other companies, such as Honda R&D and Omron, are already building and developing their tools on this platform now. “The next frontier of AI is in the physical world, and this is a once-in-a-generation opportunity for Japan,” Jensen Huang noted in the company’s official statement. He added that “Japan invented modern manufacturing. Now, it has the opportunity to reinvent it for the age of intelligent industries.”
Toyota Integrates Cars and Physical AI
The leading automotive company Toyota represents another key axis in the integration of cars and physical AI. The company uses Nvidia chips across much of its technology stack. Toyota had already committed to integrating its next-generation vehicles with Nvidia's Drive platform during the CES exhibition held in January 2025; however, the newer work extends the use of Nvidia's technology directly into its manufacturing array, where advanced simulations are used to design and plan production lines. In addition, the technology is integrated into the software running the vehicles themselves and into systems designed to read and analyze road traffic. Toyota's cars will run an advanced driver assistance system, which is capable of steering and braking but still requires the presence and intervention of a human driver. This approach is considered more conservative and cautious compared to the strategies of companies like Waymo and Tesla, which are developing and promoting autonomous driving systems that rely far less on the involvement of a human driver behind the wheel.
Why the Move Matters to Japan’s National Strategy
The significance of these moves for Japan is dramatic and backed by substantial numbers. Jensen Huang’s visit placed physical AI at the center of Japan’s industrial and national strategy, and the Tokyo government is allocating extensive financial resources to support this transition. Facing the challenge of a significant contraction in its workforce due to demographic trends, Japan aims to deploy about 10 million AI-equipped robots across 18 different industrial sectors by 2040. This ambition is backed by combined public and private investments of approximately $65 billion in the physical AI sector.
The long-term plan is even larger and more ambitious. Japan’s AI Robotics Strategy, officially released in March, aims to capture a market share of over 30% of the global AI robotics market by 2040. The Tokyo government values this global market at a total of about 20 trillion yen, equivalent to approximately $133 billion. Japan’s Ministry of Economy, Trade and Industry (METI) is funding a domestic foundation model designed to run these machines, and the joint Noetra-Nvidia data factory is where models of this massive scale—expected to reach trillions of parameters—will be trained. Japan’s bet and assumption is that the factory-floor data accumulated in the country and its broad manufacturing base will give it a massive competitive edge in the era of physical AI.
Beyond the obvious industrial considerations, there is a deep sovereign and national motive. While the United States and China lead in the development of large-scale AI systems, the Tokyo government aims to own its own data, its own independent computing power, and to significantly reduce its dependence on technological infrastructure that is not under its direct control. Huang appeared on July 16 in Tokyo alongside Japan’s Trade Minister, Ryosei Akazawa, at the government's official physical AI launch event, with Japan’s Prime Minister, Sanae Takaichi, joining the event via video call. The Takaichi administration has designated the fields of AI and semiconductors as the center of the country's economic growth plan, which aims to reach an aggregate of public and private investments of about 370 trillion yen (equivalent to approximately $2.3 trillion) by 2040. The establishment of Noetra’s AI factory—which Nvidia defines as "the world's first national AI infrastructure"—represents the country's clearest and most tangible bet in this direction. However, Japan's push for technological independence currently rests, at least at this stage, on purchasing chips manufactured in the United States.
Working Meetings with the Elite of Japanese Tech
During his short two-day visit to Tokyo, Huang made sure to meet with almost all of the most leading and influential names in the Japanese tech scene. He sat down for lunch with the CEOs of industrial giants Toyota, Fanuc, Yaskawa, Fujitsu, and Kawasaki, and met with dozens of local supply chain executives over skewers and whisky at an izakaya (traditional tavern) in the historic Kanda neighborhood. These meetings show that Huang is operating according to the same established playbook he presented just a few weeks earlier—a keynote accompanied by personal meetings in Taiwan, eating fried chicken in the markets, and signing a 50,000-GPU deal in Seoul, South Korea, last fall. This time it was Tokyo's turn, and Huang focused entirely on robots, the hardware supply chain, and the chips supporting these industries from below.