China Unicom targets 45 EFLOPS AI computing power by 2025
China Unicom reveals ambitious plan to boost intelligent computing infrastructure, aiming for 100,000-GPU clusters and 45 EFLOPS computing power by end of 2025.
China Unicom has announced a major initiative to accelerate its intelligent computing infrastructure, with plans to explore a 100,000-GPU cluster deployment and achieve 45 EFLOPS of computing power by the end of 2025. EFLOPS, a measure of computing speed, equals one quintillion floating-point operations per second.
- Strategic Vision: Chairman Chen Zhongyue revealed the plan at the company's partner conference in Shanghai, positioning China Unicom among global AI computing leaders.
- Network Upgrades: The company is advancing its Computing Power Intelligence Network with ultra-high-speed 800G and 1.2T bandwidth technologies, enabling seamless interoperability and enhanced network autonomy.
- AI Integration: By embedding AI into network operations, China Unicom aims to achieve self-configuring, self-optimizing, and self-healing capabilities, transforming network management and user experience.
Chen emphasized the role of AI agents in commercial value realization, stating they are "key to unlocking commercial value." The company's Yuanjing Wanwu AI Agent Development Platform will provide tools for model management, knowledge base construction, and workflow orchestration, enabling partners to develop AI applications with zero-code and low-code solutions.
Additionally, China Unicom launched "Tongtong Huigou," its enterprise procurement marketplace, connecting 50,000 suppliers with millions of corporate buyers. The platform has already served over 40,000 enterprises, facilitated transactions exceeding 130 billion yuan ($17.9 billion) annually, and improved order processing efficiency by 15%.
This bold strategy underscores China Unicom's commitment to leading in AI and computing infrastructure, driving innovation and efficiency across its networks and services.
Related News
AI Agents Are the New Co-Founders for Small Businesses
AI is transforming global trade, enabling small businesses to compete as micro-multinationals with lean teams and AI-powered tools.
AI Agents Pose New Cybersecurity Risks Radware Report Warns
Radware study reveals AI agents can autonomously access tools and private resources, creating new cybersecurity threats as enterprises adopt this technology.
About the Author

Dr. Lisa Kim
AI Ethics Researcher
Leading expert in AI ethics and responsible AI development with 13 years of research experience. Former member of Microsoft AI Ethics Committee, now provides consulting for multiple international AI governance organizations. Regularly contributes AI ethics articles to top-tier journals like Nature and Science.