The live broadcast on computing power and energy focused on the implementation of the integrated computing-electricity industry. Wang Shen, General Manager of China Sea Energy Storage, and Yan Bowen, Founder and Chairman of Shishi Technology, engaged in a dialogue, each drawing from their respective fields to explore the convergence and challenges of two core industries: computing power and long-duration energy storage. They systematically analyzed key topics including the practical deployment of computing-electricity synergy projects, matching long-duration storage with computing loads, and innovations in data center construction and operational models. 

From the perspective of practical implementation in the industry, both parties analyzed existing bottlenecks based on their respective domains. Yan Bowen identified two major challenges from the computing power side: first, large-scale intelligent computing clusters demand extremely high power supply stability, as fluctuations or brief power outages can cause irreversible damage to servers, chips, and storage devices; second, large-scale computing centers represent ultra-high-load electricity consumers, and deploying them within urban areas would place immense strain on the power grid. Thus, building large-scale computing clusters off-grid has become an urgent industry need. 

From the perspective of energy supply, Wang Shen identified three long-term industry bottlenecks: the cyclical and intermittent nature of renewable power generation makes it difficult to meet the real-time demand for on-demand computing; the geographical mismatch between green power in western regions and data consumption in eastern areas creates a natural conflict with the low-latency requirements of computing terminals; and computing and energy storage belong to two entirely separate technical systems, resulting in persistently high costs for cross-sector collaboration and communication. 

The most insightful part of this discussion lies in the reconfiguration of industrial economic logic and the innovation of business models. Yan Bowen, analyzing the current operational landscape of the IDC industry, pointed out that traditional data centers rely on electricity price differentials for profitability. To ensure a high availability rate of 99.99%, they require massive diesel generators and redundant UPS systems, resulting in persistently high hardware and operational costs. By leveraging long-duration energy storage systems, the online availability rate for computing services can be moderately reduced to 99.9%, significantly cutting the need for backup power equipment and substantially lowering the overall infrastructure cost for computing resources. 

Wang Shen proposed the direction for upgrading the profitability model of the computing power industry. Traditional energy storage and power companies rely solely on selling electricity for revenue, resulting in a narrow gross margin range of only 10% to 30%. However, under an integrated joint operation model, energy supply and computing power services become deeply intertwined, jointly delivering standardized computing power tokens to the AI industry. Both parties can share the value-added benefits driven by the rapid growth of the AI sector, completely breaking away from the traditional profit boundaries of single electricity or computing power sales. This represents not merely a shift in technological approach, but a higher-level evolution in industrial logic.