en.Wedoany.com Reported - On July 16, China's Ingdan Innovation launched its vertical AI Token Factory business, planning to establish distributed computing nodes at multiple locations both domestically and internationally. The company will jointly provide artificial intelligence model training and inference services through its own computing clusters and third-party computing centers at home and abroad. The company has built a vertical industry Token computing scheduling platform, with subsequent construction covering data center setup, computing cluster deployment, computing scheduling, and operation and maintenance.
This infrastructure is deployed using a "dual-track model." Part of the computing power comes from Ingdan Innovation's own clusters, while the other part connects to third-party computing centers domestically and internationally. Servers and computing resources at different locations will be scheduled by a unified platform, enabling AI model companies, embodied intelligence enterprises, and cloud service providers to call upon corresponding computing power based on task requirements. The locations, number of servers, chip configurations, and launch times of the first batch of computing nodes have not yet been disclosed.
The AI Token Factory primarily handles computing tasks during the operation of large models. After users input text, images, or device data into AI models, the server cluster must continuously complete model inference and generate output content. As model invocation increases, related business demands higher requirements for GPU servers, data center cabinets, high-speed networks, storage systems, and cluster operation and maintenance capabilities. Ingdan Innovation's entry into the Token operation segment signifies an extension of its business from AI hardware supply to computing infrastructure construction and ongoing operations.
Ingdan Innovation will provide computing partners with data center construction, computing cluster deployment, and post-operation maintenance services. The data center construction phase requires configuring server cabinets, power supply and distribution, cooling, network connections, and operation monitoring systems; after the computing cluster is put into use, it also requires equipment status monitoring, task allocation, resource utilization management, and fault handling. The company plans to connect different computing centers through a unified scheduling platform, forming a computing network from distributed devices that can provide external services.
Vertical industry applications will initially target humanoid robot and drone companies. Such devices need to process images, videos, motion trajectories, and sensor data during research and development, and conduct model training, simulation verification, and on-site inference. Ingdan Innovation plans to provide relatively lightweight Token computing services, enabling robot and drone companies to call upon computing resources based on R&D and deployment phases, without having to build complete computing clusters for each project individually.
Currently, Ingdan Innovation has accumulated over $1 billion in intended service demand, involving cloud service providers, embodied intelligence enterprises, and AI model companies. These intentions have not yet been fully converted into formal service contracts, so they cannot be directly equated with confirmed construction revenue or equipment orders, but they can serve as a basis for subsequent computing node planning and cluster expansion needs.
Subsequent physical construction progress will focus on the site selection of the first batch of distributed computing nodes, data center construction, server cluster installation, and scheduling platform integration. Only after the specific node locations, equipment scale, and launch times are announced can the actual computing power scale and infrastructure procurement demand generated by this AI Token Factory business be further assessed.
