What to watch for after Jensen Huang’s Japan visit

Nvidia chief executive Jensen Huang spent July 15 and 16 in Tokyo, holding talks with senior figures from Japan’s manufacturing, robotics, and semiconductor-related industries. Coming soon after his high-profile stop in Taiwan and following an earlier visit to South Korea, the Japan trip resulted in a wide range of new alignments across the country’s technology sector.

By the time the visit ended, Nvidia had advanced on three fronts that matter deeply to its long-term strategy: a major national AI infrastructure effort, closer cooperation with Japan’s industrial and robotics leaders, and stronger links to suppliers involved in the materials and components needed for future AI chips.

The broader significance of the trip was clear. Nvidia is pushing beyond cloud software and chatbot-style applications, aiming instead to embed AI into factories, robots, and industrial machines. Japan, with its long-established manufacturing base and robotics expertise, is emerging as a key partner in that effort.

There is also a symbolic layer to the relationship. About 30 years ago, Sega’s $5 million investment helped Nvidia through a difficult period. Today, Nvidia is once again aligning with some of Japan’s most important companies—this time around what Huang has framed as the rise of “physical AI.”

Japan’s sovereign AI push is becoming more concrete

One of the most important outcomes of Huang’s Tokyo meetings involved Noetra, a domestic initiative aimed at strengthening Japan’s own AI capabilities. The underlying goal is strategic: Japan does not want the intelligence powering its future factories, robots, and smart vehicles to rely entirely on systems developed in the US or China.

To support that ambition, around 44 Japanese companies have been assembled into a coordinated effort. Core participants include SoftBank, Sony, NEC, and Honda. Their shared objective is to build AI systems designed specifically for robotics, transportation, and industrial use inside Japan.

The Japanese government is backing the plan with substantial funding. Tokyo intends to invest up to 1 trillion yen, or roughly $6.2 billion, over five years to speed up development in what it calls physical AI. In practice, that means creating foundation models not only for digital tasks, but also for machines operating in real-world environments.

Under the emerging structure, Japan wants to maintain control over the software and model layer, while Nvidia provides the large-scale computing infrastructure required to train and deploy those systems.

Nvidia’s role centers on what it calls a Vera Rubin AI factory, a large data center based on its next-generation hardware. The facility is expected to go online in 2028 and is planned to feature 13,750 Vera CPUs and 27,500 Rubin GPUs, with total power capacity reaching 140 megawatts. Noetra is expected to lead construction and development of the site.

The project follows a phased roadmap:

• A reasoning model with strong Japanese-language performance is targeted for fiscal 2026.

• An omni-modal model capable of handling text, images, video, and audio is planned for 2028.

• A “Real-world Native AI” phase aimed at robots is scheduled for 2030, with outside developer access opening gradually.

Together, those steps reflect a larger national objective: Japan is not merely looking to adopt AI made elsewhere, but to build a domestic stack that can serve its own industrial priorities.

Nvidia is moving deeper into Japan’s factory and robotics networks

Huang’s visit also helped Nvidia strengthen its position within Japan’s manufacturing ecosystem. A broad group including Fanuc, Yaskawa, Kawasaki Heavy Industries, Fujitsu, Hitachi, NEC, Sony, SoftBank, Kubota, and robotics organization AIRoA said they plan to build on Nvidia’s Cosmos models.

Cosmos is Nvidia’s open-model platform for robotics and physical AI, introduced in May with backing from several global AI labs. During the Tokyo trip, Nvidia also unveiled Cosmos 3 Edge, a version intended to run directly inside machines using Jetson Thor chips. That matters because it shifts AI processing closer to the point of action, allowing intelligence to operate on the factory floor rather than relying only on centralized systems.

Some Japanese companies are already experimenting with shared-control systems using the platform, while others, including Honda R&D and Omron, are actively developing with Nvidia’s tools.

“The next frontier of AI is in the physical world, and this is a once-in-a-generation opportunity for Japan,” Huang said in a company statement. “Japan invented modern manufacturing. Now, it has the opportunity to reinvent it for the age of intelligent industries.”

That message is central to Nvidia’s current positioning. The company is increasingly presenting AI not just as enterprise software, but as core infrastructure for industrial automation, robotics, and the modernization of large manufacturing systems.

Toyota’s role is expanding as well

Toyota was another key focus during the visit. The automaker already relies on Nvidia technology across significant parts of its business, and at CES in January 2025 it said its next-generation vehicles would use Nvidia’s Drive platform.

The latest cooperation extends beyond vehicles themselves. It covers factory simulation tools used to design production lines, in-car software, and systems for interpreting road and traffic conditions.

Toyota’s automation strategy remains more cautious than that of some competitors. Its upcoming systems are expected to offer advanced driver-assistance features such as steering and braking support while still requiring a human driver. That places Toyota on a more measured path than companies like Waymo and Tesla, which are pursuing systems with less direct human involvement.

Why this matters for Japan’s industrial future

The Tokyo meetings were about more than a cluster of partnership announcements. They showed how deeply physical AI is now tied to Japan’s industrial planning. With the country facing a shrinking workforce, policymakers increasingly see automation and AI-driven robotics as part of the long-term solution.

Japan has set a major goal: 10 million AI-enabled robots deployed across 18 sectors by 2040. Supporting that target is an estimated $65 billion in combined public and private investment related to physical AI.

The strategy is not only about addressing labor shortages at home. In its AI Robotics Strategy released in March, Japan set out an ambition to secure more than 30% of the global AI robotics market by 2040. Tokyo values that market at around 20 trillion yen, or about $133 billion.

Japan’s Ministry of Economy, Trade and Industry is already supporting the development of a domestic foundation model for these machines. The Nvidia-backed Noetra data center is expected to become a major training base for models at very large scale, potentially reaching into the trillions of parameters.

At the center of all this is a question of sovereignty. As the US and China continue to pull ahead in large-scale AI, Japan is trying to ensure it has its own access to computing infrastructure, its own model foundations, and more control over strategic AI systems.

That theme was visible during the government’s physical-AI launch event on July 16, when Huang appeared alongside trade minister Ryosei Akazawa, while Prime Minister Sanae Takaichi joined by video. The administration has made AI and semiconductors key pillars of an economic growth strategy targeting 370 trillion yen, or about $2.3 trillion, in combined public and private investment by 2040.

Noetra’s planned AI factory—described by Nvidia as “the world’s first national AI infrastructure”—offers one of the strongest examples yet of that strategy moving from rhetoric to execution. At the same time, it highlights a tension in Japan’s position: the country wants greater independence in AI, but for now its plans still depend heavily on US chip technology.

Huang’s approach followed a familiar regional pattern

Over the course of the two-day trip, Huang met with many of the most influential leaders in Japanese industry. His schedule included a lunch discussion with the chief executives of Toyota, Fanuc, Yaskawa, Fujitsu, and Kawasaki, as well as meetings with dozens of supply-chain executives at a Kanda izakaya.

The method echoed the relationship-building playbook Nvidia used in Taiwan just weeks earlier and in South Korea during its push for deals last fall. In Japan, though, the priorities were especially focused: robotics, industrial automation, semiconductor inputs, and the computing systems underneath them.

That is why Huang’s Tokyo visit deserves attention beyond the immediate headlines. It was not simply a ceremonial appearance or a routine executive stop. It was a coordinated effort to tie Nvidia more closely to Japan’s industrial transition—at a time when the country is trying to build a more self-directed position in the next era of AI.