Japan launches the domestic AI project “Noetra”: Nvidia supplies 27.5k Rubin GPUs, with 44 enterprise groups including Sony and SoftBank participating

BlockBeats message: On July 17, Japan’s domestic multimodal foundation model R&D project “Noetra” officially launched on July 16. Its core enterprises are Sony Group, SoftBank, NEC, and Honda. A total of 44 companies and organizations will contribute funding and participate, covering a broad range of industries such as manufacturing, finance, logistics, and communications. Engineers from organizations including the National Institute of Advanced Industrial Science and Technology (AIST) and Preferred Networks will also join the development. The project falls under the “Multimodal foundation model development program for AI robots and physical AI” being advanced by Japan’s Ministry of Economy, Trade and Industry (which NVIDIA officially refers to as the FRONTia Project). The initiative aims to build Japan-based foundation models for physical AI scenarios such as manufacturing sites and robots, rather than merely adding a Japanese conversational AI.

On the hardware side, Noetra will collaborate with NVIDIA to build a computing platform equipped with about 27,500 of the latest Rubin GPUs and 13,750 Vera CPUs. It will adopt the NVIDIA Vera Rubin NVL72 rack and DSX platform architecture. The design provides an electrical power capacity of 140 megawatts. Construction is planned to begin in April 2027, with production scheduled to start in June 2028, at which point it will become Japan’s largest AI compute infrastructure.

The R&D roadmap will be carried out in three phases: starting in the 2026 fiscal year, developing inference foundation models centered on AI agents and natural language processing; in the 2028 fiscal year, achieving full multimodal foundation models with seamless fusion of text, images, video, and audio; and in the 2030 fiscal year, enabling real-world native AI that understands space and physical attributes, ultimately to be applied in fields including manufacturing, logistics, healthcare, and communications.