The 2025 China Computing Power Conference recently introduced that the scale and level of China's computing infrastructure have significantly improved, and the computing power industry has entered a new stage of high-quality and large-scale development. Computing power is regarded as the core productivity of the digital economy era. Only by rooting computing power in industries and deeply integrating it with the real economy can it unleash its actual value and truly transform into productivity. What is computing power? Computing power, also known as computing capability, is a new type of productivity that integrates information computing power, network transportation power, and data storage power. It mainly provides services to society through computing infrastructure. Simply put, computing power is like a super brain capable of processing massive amounts of data. We rely on this invisible super brain to smoothly shop online, navigate routes, and communicate with AI assistants. There is a professional indicator called FLOPS to measure the strength of computing power, which is the number of floating-point operations per second. At present, computing power can be roughly divided into three categories: basic computing power, intelligent computing power, and supercomputing power. The strongest computing power among them is supercomputing, which can complete 10 billion billion operations per second, equivalent to the computational workload of 8 billion people pressing keys continuously for nearly 40 years worldwide. China's computing power industry is in the leading tier globally. In terms of scale, China's total computing power ranks second in the world, and the upstream and downstream industrial chains are gradually improving. It has a solid foundation in using standard racks from computing power centers and the number of 400G ports on trunk lines. From a structural perspective, the rapid development of intelligent computing power indicates that the quality of computing power in China is constantly improving. Basic computing power, intelligent computing power, and supercomputing power support basic general computing, artificial intelligence computing, and scientific engineering computing. Previously, basic computing power accounted for the majority. Nowadays, the proportion of intelligent computing power continues to increase, which means that the ability to handle complex tasks is stronger, and structural optimization will strongly support digital transformation. The key to transforming computing power into productivity lies in its deep integration with the real economy. Computing power is often compared to a new type of electricity, but unlike electricity, computing power does not generate value simply by being connected to a circuit. If computing power is disconnected from specific industrial scenarios, it will become an "idle resource" and even cause resource waste. In recent years, the utilization rate of many computing power centers has been very low, with machines idle "basking in the sun during the day and counting stars at night". One important reason is that before construction, there was no in-depth research on local industrial demand, and only focused on current policy dividends and short-term benefits blindly following the trend. After construction, there was a lack of matching industrial demand, resulting in idle computing power and frequent waste of resources. Industrial demand guides the direction of computing power development. Breaking away from industrial demand, computing power construction is aimless. Industrial applications are the touchstone of computing power value, providing clear market direction and landing scenarios for its evolution and upgrading. For example, the traditional manufacturing industry has low production efficiency and high energy consumption, which has created a demand for edge computing power and intelligent computing power. By deploying edge computing nodes on the production line, real-time processing of massive data collected by sensors, and using AI computing power to optimize production processes, intelligent manufacturing can be achieved; Traditional agriculture needs computing power support such as satellite remote sensing and big data to break away from relying solely on the weather for food. This enables early warning of pests and diseases, precise fertilization, and more. The core driving force for the upgrading and iteration of computing power technology comes from these practical industry pain points. The upgrading of computing power can not only promote industrial upgrading, but also break down industrial boundaries and stimulate new industries. As a core production factor, computing power can break through existing technological bottlenecks through deep processing of data, continuously unlock previously difficult application scenarios, and promote new industries from concept to reality. On the one hand, the upgrading of computing power has driven the formation of new industries in the upstream and downstream of the computing power industry chain, such as the explosive growth of AI computing power, which has given rise to segmented industries such as AI model training services and computing power leasing; On the other hand, the deep integration of computing power and traditional industries has brought new industries such as smart agriculture, industrial Internet, and digital cultural innovation. Computing power will eventually bloom and bear fruit in the industrial soil. At present, China's computing power industry is leading in scale and technology, and has a solid foundation to support the transformation of the real economy. Based on industrial demand and leveraging the advantages of China's large market and multiple scenarios, we will promote continuous iteration and optimization of computing technology, allowing computing power to take root and truly transform into a strong engine for high-quality development of the Chinese economy. (New Society)