Japan targets the launch of a vast datacenter utilizing NVIDIA's Vera Rubin GPUs by June 2028.
This initiative highlights Japan's ambition to strengthen its position in the global AI infrastructure arena, fueled by the demand for advanced computing capacity for next-gen AI applications.
The upcoming datacenter will be supported by NVIDIA's next-generation Vera Rubin architecture, which is set to replace the current Blackwell design. Each NVL72 rack will contain 72 Rubin GPUs alongside Vera CPUs and NVLink 6 interconnects, with volume shipments slated to commence in fall 2026.
In addition to this project, South Korea’s SK Group has unveiled plans for an “AI Factory” datacenter in Japan, expected to be operational between 2028 and 2029. Furthermore, NVIDIA is reportedly engaging with Mitsubishi Heavy Industries to create specialized cooling and power systems tailored for AI datacenters.
Hitachi is also contributing to the ecosystem with a memorandum signed with Mitsui OSK Lines aimed at developing a floating datacenter in Japan. As the demand for AI hardware surges, this datacenter initiative opens multiple investment opportunities, positioning NVIDIA as a key beneficiary.
However, the timely execution of the project is contingent on NVIDIA meeting its shipping schedules, maintaining the construction timeline, and ensuring adequate power grid capacity.
This material is informational and not financial advice.
