NVIDIA is working with Noetra to build an NVIDIA Vera Rubin AI factory in Japan, a project backed by Japan’s Ministry of Economy, Trade and Industry (METI) and tied to the country’s FRONTia program for multimodal foundation models aimed at physical AI. The build centers on an NVIDIA Vera Rubin platform, with NVIDIA calling it the world’s first national AI infrastructure for physical AI.
The AI factory is described as using NVIDIA Vera Rubin NVL72 racks on the NVIDIA DSX platform, with NVIDIA Spectrum-X Ethernet for cluster connectivity and scaling. NVIDIA also lists NVIDIA BlueField DPUs as part of the architecture. Capacity for the site is specified at 140 MW of data center capacity.
On the compute side, the configuration calls for 13,750 NVIDIA Vera CPUs and 27,500 NVIDIA Rubin GPUs. That’s a clear signal this is being designed as a very large, centralized training environment rather than a smaller, distributed approach, and the 140 MW figure puts power delivery, electrical protection, and heat rejection at the center of the engineering problem from day one.
Noetra’s role includes establishing the AI factory and developing multimodal foundation models. NVIDIA says pretrained weights from Noetra’s multimodal models will be made broadly available to domestic model developers and enterprises, alongside NVIDIA software and models including NVIDIA Nemotron, NVIDIA Cosmos, NVIDIA Isaac GR00T open models, and NVIDIA NeMo libraries.
METI’s FRONTia Project is referred to as “Development of Multimodal Foundation Models with a View to AI Robotics and Physical AI.” The stated goal is to use Japan’s industrial data and manufacturing expertise to develop “highly reliable multimodal foundation models” for physical AI applications, including AI agents, digital twins, and robotics.
Jensen Huang, founder and CEO of NVIDIA, said, “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, “By fostering collaboration between Japan and leading global innovators—including NVIDIA—and leveraging 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.”
Hironobu Tamba, CEO of Noetra, said, “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.”
NVIDIA describes DSX as a reference design and platform for AI factories, intended to accelerate time to production and improve token throughput per megawatt. It also states the AI factory architecture is intended to reduce token costs and support “frontier AI training,” including training “trillion-parameter-scale” models as the build expands.
Source: NVIDIA












