Japan's Ministry of Economy, Trade and Industry (METI) has partnered with NVIDIA and cloud provider Noetra Corp. to build the world's first national AI infrastructure dedicated to physical AI. The 140-megawatt AI factory, announced on July 16, 2026, will house 13,750 NVIDIA Vera CPUs and 27,500 NVIDIA Rubin GPUs, supplying the computing foundation for Japan's industrial AI strategy across manufacturing, logistics, healthcare, and robotics.
The new facility, established by Noetra, will use NVIDIA Vera Rubin NVL72 racks on the NVIDIA DSX platform, connected and scaled with NVIDIA Spectrum-X Ethernet networking. It will enable the development of open multimodal foundation models that drive AI agents, digital twins, robotics, and other physical AI applications.
"Japan invented modern manufacturing. Now, it is building the AI factories that will power the next industrial revolution," said Jensen Huang, founder and CEO of NVIDIA. "NVIDIA is honored to partner with Japan and its industrial leaders to build the AI infrastructure that will power the country's industries, its economy and a new generation of innovation."
The FRONTia Project and open foundation models
The AI factory will serve as the computing backbone for Japan's FRONTia Project, which METI launched to develop highly reliable multimodal foundation models for physical AI. The project brings together the country's manufacturing expertise, real-world industrial data, and global technology partners. The initiative reflects a broader shift toward AI for Government, where national IT infrastructure is treated as a strategic asset for industrial competitiveness.
Noetra will make the pretrained weights of its multimodal models openly available to domestic developers and enterprises, alongside software such as NVIDIA Nemotron, NVIDIA Cosmos, and NVIDIA Isaac GR00T open models. This approach aims to accelerate the creation of agentic AI and physical AI applications across Japanese industries.
"Japan has launched the FRONTia Project, which will serve as the core of the country's physical AI ecosystem," said Ryosei Akazawa, Japan's Minister of Economy, Trade and Industry. "By fostering collaboration between Japan and leading global innovators - including NVIDIA - and using Japan's strengths, such as its onsite expertise and manufacturing technology infrastructure, we will build highly reliable multimodal foundation models and contribute to solving global social challenges."
Vera Rubin architecture: what's inside the AI factory
The AI factory's design combines NVIDIA Vera Rubin GPUs and CPUs with the DSX reference architecture, BlueField DPUs, and tightly codesigned silicon, systems, and software. The Spectrum-X Ethernet networking platform ties the components together. NVIDIA says this integration delivers breakthrough AI performance, lower token costs, and the scale needed for frontier AI training.
The 140-megawatt data center will support training of trillion-parameter-scale AI models as it expands, giving organizations across Japan access to one of the world's most advanced AI environments. The DSX platform helps infrastructure builders reach production faster, increase token throughput per megawatt, and operate with greater reliability.
Japan's push for AI robotics leadership
Japan's AI Robotics Strategy, released in March, sets a target to capture more than 30% of the global AI robotics market by 2040, a $133 billion opportunity. METI's multimodal foundation model program for robotics and physical AI is a central part of that policy. The new AI factory directly supports these goals by providing the computing needed to train, test, and deploy physical AI at scale.
"Bringing physical AI into the real world requires enormous computing, data and foundational technologies - challenges no single company can solve alone," said Hironobu Tamba, CEO of Noetra. "Together with partners across Japan and around the world, Noetra will advance Japan-developed multimodal foundation models and accelerate the deployment of physical AI across Japanese industries by broadly sharing the results of our research."
Why this matters for government professionals
This initiative is a concrete example of how a national government can partner with private sector technology leaders to build sovereign AI infrastructure. It shows a model for public investment that targets open models and domestic developer access, rather than relying solely on proprietary cloud services. The focus on physical AI - robotics, manufacturing, logistics - ties directly to industrial policy and economic competitiveness.
Government professionals involved in AI strategy, digital infrastructure, or industrial policy can draw lessons from METI's approach: setting clear market-share targets, structuring public-private partnerships, and open-sourcing model weights to speed up domestic adoption. For those shaping similar programs, the AI Learning Path for Policy Makers offers a structured way to build the technical and policy knowledge needed to evaluate such national AI investments.
Your membership also unlocks:
