Kanghui Shares: Subsidiary Signs Computing Power Service Contract, Total Amount Expected to Be 415 Million to 679 Million Yuan 康惠股份：子公司签署算力服务合同，预计总金额4.15亿元至6.79亿元

Summary

Kanghui Shares’ wholly-owned subsidiary signed a computing power service contract with Company D, involving the procurement, deployment, and full-lifecycle operation and maintenance of high-performance servers.
The total contract value ranges from 415 million RMB to 679 million RMB, with a cooperation term of five years (including an initial three-year period and a two-year renewal period).
The service model includes equipment networking, stress testing and joint debugging, and monthly billing for computing power services, reflecting the commercialization of computing infrastructure operations.

Deep Analysis

TL;DR

• Kanghui Shares’ wholly-owned subsidiary signed a computing power service contract with Company D, involving the procurement, deployment, and full-lifecycle operation and maintenance of high-performance servers.
• The total contract value ranges from 415 million RMB to 679 million RMB, with a cooperation term of five years (including an initial three-year period and a two-year renewal period).
• The service model includes equipment networking, stress testing and joint debugging, and monthly billing for computing power services, reflecting the commercialization of computing infrastructure operations.

Why It Matters

This case illustrates the specific path for traditional enterprises to transform into computing power service operators, providing a real-world transaction reference for business models in the computing power rental market. For investors focused on the AI infrastructure industry chain, such large-scale long-term contracts help verify the sustainability of computing demand and the revenue growth potential of related companies.

Technical Analysis

• Hardware Deployment: Responsible for procuring and deploying "high-performance computing servers." While specific chip models were not disclosed, they must meet the particular technical indicators required by Client D.
• Networking and Debugging: Execution of engineering tasks such as equipment networking, stress testing, and joint debugging to ensure cluster stability and performance compliance, which is a critical step in delivering computing power services.
• Operation and Maintenance System: Provides full-lifecycle O&M services, covering equipment maintenance, fault handling, and performance monitoring from acceptance through the end of the contract, ensuring the continuity of computing power services.

Industry Insights

• Normalization of Computing Power Rental: The emergence of large-scale computing contracts indicates that computing power has shifted from simple hardware sales to a long-term "hardware + service" operational model, with a clear trend toward SaaS/IaaS integration.
• Surge in Vertical Demand: Client D’s large-scale computing power procurement suggests that rigid demand for high-quality computing power in AI application layers or specific industries is being released.
• Restructuring of Industry Chain Value: Companies capable of integrated deployment and O&M will occupy a more important position in the computing power value chain. Pure hardware manufacturers need to extend into services to enhance their competitiveness.

TL;DR

• 康惠股份全资子公司与D公司签署算力服务合同，涉及高性能服务器采购、部署及全周期运维。
• 合同总金额区间为4.15亿元至6.79亿元，合作期限5年（含前3年初始期及后2年续约期）。
• 服务模式包括设备组网、压测联调及按月收取算力服务费，体现算力基础设施运营的商业化落地。

为什么值得看

该案例展示了传统企业向算力服务运营商转型的具体路径，为算力租赁市场的商业模式提供了真实交易参考。对于关注AI基础设施产业链的投资者而言，此类大额长期合同有助于验证算力需求的持续性及相关企业的营收增长潜力。

技术解析

• 硬件部署：负责采购并部署“高性能算力服务器”，具体芯片型号未披露，但需满足客户D公司的特定技术指标。
• 网络与调试：执行设备组网、压测联调等工程化工作，确保集群稳定性与性能达标，这是算力服务交付的关键环节。
• 运维体系：提供全周期运维服务，涵盖从验收后到合同结束期间的设备维护、故障处理及性能监控，保障算力服务的连续性。

行业启示

• 算力租赁常态化：大型算力合同的出现表明，算力已从单纯的硬件销售转向“硬件+服务”的长期运营模式，SaaS/IaaS化趋势明显。
• 垂直领域需求爆发：D公司作为特定客户，其大规模算力采购暗示了AI应用层或特定行业对高质量算力的刚性需求正在释放。
• 产业链价值重构：具备集成部署和运维能力的企业将在算力价值链中占据更重要位置，单纯硬件制造商需向服务延伸以提升竞争力。