During a two‑day visit to Tokyo, Nvidia CEO Jensen Huang sealed a series of agreements that stitch the company deeply into Japan’s industrial future. The deals span a government‑backed sovereign AI effort, a wide coalition of robotics and manufacturing firms, and supply partnerships to power next‑generation AI chips. Together they position Nvidia at the heart of Tokyo’s push to put AI into factories, vehicles and robots.
Three headline projects
Huang’s itinerary produced three headline initiatives that illustrate how Nvidia intends to anchor itself in Japan’s physical‑AI strategy.
First is Noetra, Tokyo’s domestic AI consortium. The government assembled roughly 44 Japanese firms, including SoftBank, Sony, NEC and Honda, to build foundation models aimed at robots, vehicles and factory automation. Tokyo has pledged up to 1 trillion yen ($6.2 billion) over five years to the effort, which is explicitly designed to give Japan its own software brain for machines rather than rely on foreign providers. Nvidia will supply the hardware backbone: the company is constructing what it calls a Vera Rubin AI factory — a large data center slated to begin operations in 2028 and packed with next‑generation chips. The facility is expected to include 13,750 Vera CPUs and 27,500 Rubin GPUs and to deliver 140 megawatts of power.
Noetra’s roadmap runs in stages. The first phase targets a reasoning model with strong Japanese language capabilities beginning in fiscal 2026. By 2028 Tokyo expects an omni‑modal model that handles text, images, video and audio. The final stage, slated for 2030, is “Real‑world Native AI” tailored to operate robots; Noetra plans phased releases to outside developers.
Second is a broad robotics coalition rallying behind Nvidia’s Cosmos models. Major industrial names — Fanuc, Yaskawa, Kawasaki Heavy, Fujitsu, Hitachi, NEC, Sony, SoftBank, Kubota and robotics group AIRoA — have said they will build on Cosmos. Nvidia unveiled Cosmos 3 Edge in Tokyo, a version that can run inside machines on Jetson Thor chips, enabling models to operate at the edge within robots and manufacturing equipment.
Third, Toyota deepened its existing relationship with Nvidia. The automaker already uses Nvidia chips across much of its stack and committed next‑generation vehicle development to Nvidia’s Drive platform in January 2025. The new work extends Nvidia’s reach into Toyota’s manufacturing simulations, vehicle software and systems that process traffic information. Toyota’s approach will focus on advanced driver assistance that still requires a human driver, rather than fully driverless systems.
Why Japan is betting on factory‑floor AI
The deals reflect a national strategy that treats physical AI as central to Japan’s economic future. Facing a declining workforce, Tokyo has set an ambitious target of deploying 10 million AI‑equipped robots across 18 sectors by 2040, backed by roughly $65 billion in public and private physical‑AI investment. The government’s AI Robotics Strategy, published in March, aims for Japan to capture more than 30% of the global AI robotics market by 2040 — a market Tokyo values at about ¥20 trillion (around $133 billion).
METI is funding a domestic foundation model intended to run machines, and Noetra’s Nvidia‑powered data center is positioned as the training hub for models at trillion‑parameter scale. The underlying bet is that Japan’s manufacturing data and industrial base can be turned into a competitive advantage for real‑world AI deployed in factories and on roads.
The sovereign and supply angles
Beyond the industrial case lies a sovereign calculation. Japan wants to reduce dependence on foreign AI infrastructure as the U.S. and China accelerate large‑scale AI development. Establishing domestic compute, data stewardship and software stacks is a priority for Tokyo’s current administration, which has made AI and semiconductors a pillar of a long‑range growth plan that targets extensive public and private investment by 2040.
Despite the emphasis on homegrown models, Japan’s plan still relies on American silicon. Nvidia’s data center and chip commitments illustrate that the country will depend on U.S. hardware even as it seeks software autonomy. Huang’s appearance alongside trade minister Ryosei Akazawa at the government’s physical‑AI launch — with Prime Minister Sanae Takaichi joining by video — underscored the high level of political backing for the effort.
Huang’s outreach and the broader strategy
The Tokyo visit continued a pattern of direct, relationship‑driven diplomacy by Huang. Over two days he met CEOs and supply‑chain leaders from Toyota, Fanuc, Yaskawa, Fujitsu and Kawasaki, and held informal gatherings with dozens of industry executives. The approach mirrors recent stops in Taiwan and South Korea, where Huang has combined technical presentations with social engagements to win commitments.
For Japan’s industrial giants, the collaborations offer a path to deploy AI at scale within production lines and robots. For Nvidia, the deals secure a pivotal role in the nascent physical‑AI market, from chips and edge devices to the data centers that will train national models. That relationship mirrors an earlier moment in the companies’ histories — decades ago a $5 million investment from Sega helped buoy Nvidia — but today the partnership is aimed at building the machines that will power the next era of intelligent industry.
The Tokyo agreements make clear that Japan intends to be an active builder of physical AI, not just a consumer. Whether the combination of domestic models, industrial know‑how and U.S. compute will deliver the promised gains remains to be seen, but the country’s strategy now centers on factory floors, robots and the chips that run them.
Source: TechCrunch AI