Japan's Ministry of Economy, Trade and Industry (METI), industrial leaders, and NVIDIA are building the world's first national AI infrastructure purpose-built for physical AI. The AI factory, operated by Noetra Corp., will deploy 13,750 NVIDIA Vera CPUs and 27,500 NVIDIA Rubin GPUs to deliver 140 megawatts of data center capacity. The computing power will support AI agents, digital twins, robotics, and other physical AI applications across manufacturing, logistics, healthcare, and telecommunications.
The facility will serve as the computing foundation for Japan's FRONTia Project, a government-backed program to develop open multimodal foundation models. These models, along with software such as NVIDIA Nemotron, Cosmos, and Isaac GR00T, will be shared broadly with domestic developers and enterprises to speed up the creation of agentic AI and physical AI systems.
"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."
Ryosei Akazawa, Japan's Minister of Economy, Trade and Industry, said the FRONTia Project "will serve as the core of the country's physical AI ecosystem." He added that collaboration with global innovators and Japan's manufacturing expertise will help "build highly reliable multimodal foundation models and contribute to solving global social challenges."
Architecture and Scale
The AI factory uses the NVIDIA Vera Rubin DSX reference design, with NVL72 racks and NVIDIA Spectrum-X Ethernet networking to connect and scale the infrastructure. NVIDIA BlueField DPUs and tightly codesigned silicon, systems, and software are designed to deliver higher token throughput per megawatt and lower token costs for frontier AI training.
Noetra CEO Hironobu Tamba said the complexity of bringing physical AI into real-world settings "requires enormous computing, data and foundational technologies - challenges no single company can solve alone." He said the initiative will advance Japan-developed multimodal models and "accelerate the deployment of physical AI across Japanese industries by broadly sharing the results of our research."
Japan's AI Robotics Ambitions
Japan's AI Robotics Strategy, released in March 2026, targets capturing more than 30% of the global AI robotics market by 2040, an opportunity estimated at $133 billion. METI's push for a multimodal foundation model program aligns with that goal, and the AI factory will eventually support training trillion-parameter-scale models, giving organizations across the country access to some of the world's most advanced AI training environments.
Why this matters for Government
Japan's approach demonstrates a model for public investment in shared AI infrastructure that coordinates industrial policy, attracts global technology partners, and sets national standards for model development. The open release of foundation model weights raises practical considerations around procurement, data governance, and interagency collaboration that policy makers in other countries will need to address as similar national AI initiatives take shape. Government professionals tracking these developments can find relevant analysis through AI for Government coverage, and those directly involved in shaping national AI strategy may benefit from the structured guidance in the AI Learning Path for Policy Makers.
Your membership also unlocks:
