Partnership between JERA, Dell, and RHAELM plans $15 billion AI megacomplex in Japan

A partnership between Japanese power generator JERA, equipment manufacturer Dell Technologies, and UK developer RHAELM was announced on Thursday (1) to build an artificial intelligence data center in Chiba, near Tokyo. The initial project involves a $15 billion investment and a 400-megawatt power capacity, with the goal of integrating power generation and computing in the same location.

The investment could be part of a larger expansion of the sector in the country, valued at up to $140 billion, according to the Financial Times. This figure does not represent a public investment but rather a projection of corporate spending over several phases.

Energy generated on-site

The complex will be located next to JERA’s Chiba thermal power plant. Electricity will be supplied directly by the company’s plant in a configuration known as behind-the-meter, which reduces reliance on conventional grid connection timelines.

“We can deliver a 400 MW facility years ahead of a conventional grid connection schedule,” said Bradd Lewis, CEO of RHAELM, in the partnership announcement.

The strategy addresses the growing energy consumption of data centers, which, according to the International Energy Agency, could reach 950 terawatt-hours by 2030—nearly double the current level.

Roles of each company and financing

JERA contributes land and energy; Dell provides server and computing infrastructure; and RHAELM leads the project’s development. Apollo Global Management is expected to participate as an investor and financier of RHAELM.

Timeline and future prospects

Operations are scheduled to begin in 2028, with full capacity expected by 2029. The companies plan to use the project as a model for future facilities, replicating the integration of generation, cooling, and computing in other JERA units across Japan, with the goal of reaching several gigawatts in the 2030s.

If realized, the Japanese expansion could become one of the world’s largest investments in AI infrastructure, reflecting the global race for computing capacity and reliable energy.