Dell, Apollo, and Jera have unveiled plans to invest as much as $140 billion developing AI-focused data centers and associated gas power plants across Japan and Asia. The initiative targets rapid deployment of up to 4GW of AI compute within five years, starting with a $15 billion 400MW facility near Tokyo slated for 2028.
- Up to 4GW of AI data centers matched with dedicated gas power plants
- Initial $15B Tokyo-area facility to begin operations by 2028
- Standardized, scalable infrastructure model planned for repeat deployments
Infrastructure signal
This new initiative promises a substantial expansion of AI data center infrastructure, combining advanced computing capacity with dedicated gas-fired power generation. By developing as much as 4GW of AI compute resources over five years, the project addresses critical infrastructure needs for electricity-intensive AI workloads and reduces dependency on national power grids by operating independent power facilities.
The partnership’s upfront $15 billion investment for the initial 400MW data center near Tokyo highlights the capitalization and scale necessary to support cutting-edge AI workloads in Asia. Standardization of the data center design and integrated power infrastructure is envisioned to promote faster deployment and replication across multiple sites, potentially expanding beyond Japan to the broader Asian region.
Developer impact
Developers working in AI and other compute-heavy domains in Japan and Asia can expect significant improvements in infrastructure reliability and availability, with dedicated power ensuring stable, high-performance environments. This reduces risks of interruptions and latency caused by grid fluctuations and supports large-scale AI model training and inferencing workflows.
The rapid deployment schedule and standardized facility approach promise more predictable resource availability and potential API or platform service enhancements tied to these new data centers. This could foster better integration with cloud-native development workflows, accelerating AI product development and deployment timelines for regional teams.
What teams should watch
Teams responsible for infrastructure strategy, cloud cost management, and platform operations should track progress on the initial Tokyo project slated for 2028, as its performance and operational metrics will set benchmarks for scaling AI infrastructure across Asia. It will be important to observe how the independent electricity generation impacts operational costs and reliability compared to traditional grid-dependent data centers.
Developer operations and product teams should prepare for evolving API interfaces and deployment models that may arise from this tightly integrated compute and power infrastructure model. Additionally, database and observability teams need to focus on handling increased data throughput and monitoring capabilities suitable for large-scale AI workloads enabled by these new robust data centers.