Japan Lands $15B AI Data Center as Dell and Apollo Back 400MW Project

The companies said in their official memorandum of understanding that the project will combine power generation, cooling and standardized AI compute infrastructure at JERA’s Chiba Thermal Power Station. Apollo Global Management intends to participate as RHAELM’s strategic investment and financing partner.

The bigger angle is not simply the size of the campus. The partners want Chiba to become a repeatable model for national-scale AI infrastructure, potentially expanding across other JERA power sites during the 2030s.

400MW Project Bypasses a Major AI Bottleneck

Power availability has become one of the biggest obstacles to building large AI data centers.

The Chiba project is designed as a behind-the-meter facility, meaning it can connect directly to JERA’s existing generation infrastructure rather than waiting years for a conventional grid connection.

That gives the project an advantage at a time when electricity is becoming one of AI’s biggest infrastructure constraints.

JERA will provide the land and round-the-clock power, while Dell will supply standardized rack-scale AI infrastructure. RHAELM will handle development and delivery.

Operations are targeted to begin around 2028, with Reuters reporting that full capacity is expected by 2029.

AI Infrastructure Is Becoming an Energy Trade

The structure shows how far the AI investment story has moved beyond GPUs.

Modern data centers require enormous amounts of capital for power generation, cooling, networking, servers and construction. That has created a parallel financing boom, with private credit, bonds and infrastructure investors funding trillions of dollars of AI expansion.

Apollo’s involvement is therefore as important as Dell’s.

The project also strengthens Dell’s exposure to AI infrastructure after its AI server backlog climbed to $95 billion.