With a total planned investment of up to RMB5 billion, the project is designed to support Chinese government's national initiatives of "East-to-West Computing Resource Transfer" and "Xinjiang-to-Chongqing Computing Resource Transfer". The project aims to build an intelligent computing infrastructure network covering key regions, deliver rapidly deployable standardized computing units and enhance regional computing supply capacity and utilization efficiency.

Against the backdrop of the Chinese government's 15th Five-Year Plan's push to transform the national computing infrastructure system from "construction-focused" to "operation and secure and controllable-focused," the Stars Project adopts a forward-looking three-layer collaborative architecture: "dual-core + multi-level edge + unified platform". This architecture will directly address the challenges of latency, bandwidth costs, and data compliance bottlenecks when deploying AI applications to end-user scenarios such as industrial and urban governance.

"Dual-Core + Multi-level Edge Nodes": Building a Nationally Integrated Agile Computing Network

The Stars Project is planned with a construction period of 60 months and will be rolled out in five phases, from pilot validation to large-scale implementation. Its core infrastructure will include:

  • Dual-Core Intelligent Computing Centers (Centralized Training & Inference Hub): The computing centers will be located in Yinchuan, Ningxia, with a planned deployment of 512 high-performance servers, and Yiwu, Xinjiang, with a planned deployment of 256 high-performance servers. The dual-core centers will operate in coordination to provide robust centralized training, inference, and massive data processing capabilities.



     
  • 50–100 Multi-level Edge Nodes (Low-Latency Outposts): These nodes will be deployed in containerized, modular form, enabling "on-site installation, plug-and-play, elastic expansion". Each node is equipped with 10–16 edge servers for low-latency inference and localized data processing, fully suitable for multi-site replication needs.



     
  • Integrated Command and Dispatch Platform: A unified national brain platform will be responsible for management, scheduling, billing, operation and maintenance, and security of the entire network's computing resources.



     

First 4000P Edge Computing Node Deployed, Redefining Agile Computing

To address latency and data residency challenges of traditional centralized intelligent computing centers, the Stars Project will promote edge computing when responding to end-user scenarios. This solution will adopt a containerized, modular deployment method with significant advantages in plug-and-play and elastic expansion.

The implementation of the Stars Project and investment, cooperation and other arrangements contemplated therein will be subject to further negotiation and execution of definitive agreements among the relevant parties.

In early 2026, Huazhi Future delivered a dedicated briefing on the Stars Project to the Yuzhong District government of Chongqing, and received high recognition and support from the government. In April 2026, it officially deployed its first high-performance edge computing node in Lengshui Town, Shizhu County, Chongqing, with a scale of 4000 Petaflops ("P"), while actively coordinating green power direct supply to create a zero-carbon computing benchmark.