Summary of Key Points
Recently, the explosive growth of large AI models and data centers has led to a surge in domestic demand for computing power and electricity. Taking Guangdong as an example, the total electricity consumption by data centers in the first half of 2026 increased by 21.15% year-on-year. Among them, Shaoguan saw a particularly high increase in electricity consumption of 409% due to hosting the only national computing power hub cluster in the Guangdong-Hong Kong-Macao Greater Bay Area. To address this trend, "computing-power synergy" (using electricity to support computing power and using computing power to optimize electricity usage) has become a crucial solution. Nationally, although the current proportion of electricity consumed by computing power is small (1.6% in 2025), it is expected to grow rapidly in the medium to long term, accounting for 6% of the total social electricity consumption by 2030. The government is promoting computing-power synergy to resolve issues such as the imbalance of resources between the eastern and western regions (where the east has high demand but limited energy, while the west has abundant energy but low demand).
1. Why Has Shaoguan Become the Leader in Guangdong's Computing-Power Consumption Growth?
The reason for Shaoguan's significant increase in electricity consumption (409%) is that it hosts the only national computing power hub data center cluster in the Guangdong-Hong Kong-Macao Greater Bay Area. To meet the electricity needs of these data centers, the Guangdong power grid has been specifically planning its infrastructure since 2022. As of June 2026, 740 million yuan have been invested to build three substations with a voltage capacity of over 110 kV, providing a total power supply capacity of 840 megavolt-amps. In simple terms, for every additional 10,000 server racks installed, the electricity demand increases by 25,000 kilowatts. Shaoguan is rapidly expanding its data center capacity, which is naturally resulting in a sharp rise in electricity consumption.
2. How Does Computing-Power Synergy Help Both Power Grids and Computing Facilities?
Computing-power synergy aims to optimize the use of both power and computing resources in two steps:
- Power Supporting Computing: The power grid makes advance arrangements to ensure that data centers have a stable supply of electricity. For example, Shaoguan plans its substations in a way that aligns with the layout of server cabinets, ensuring that the power grid and computing facilities develop in tandem and preventing situations where computing power is available but not used.
- Computing Optimizing Power: Data centers can play a role in helping the power grid. For instance, during peak electricity consumption times, non-urgent tasks (such as data backups) can be scheduled for off-peak periods at night. They can also connect to virtual power plant platforms to adjust their energy usage on a minute-by-minute basis, which not only reduces their own electricity costs but also helps the power grid manage demand peaks and troughs. Guangdong has already implemented the first domestic transaction involving a data center-based virtual power plant, with computing clusters in Shaoguan, Guangzhou, and Zhanjiang participating in this initiative.
3. National Electricity Consumption for Computing Power: Low Impact Now, but Growing Pressure in the Future
Currently, the impact of electricity consumption for computing power on the overall power supply and demand is limited. In 2025, the national electricity consumption for computing power was approximately 170 billion kilowatt-hours, accounting for only 1.6% of the total. The main peak this summer was due to air conditioning usage. However, the long-term growth potential cannot be underestimated. According to predictions by the China Electricity Council, the annual increase in electricity consumption for computing power is expected to exceed 100 billion kilowatt-hours during the "14th Five-Year Plan" period, reaching 800 billion kilowatt-hours by 2030 (6% of the total). Computing power will become an important driver of future electricity consumption.
4. How Does Computing-Power Synergy Solve the Resource Mismatch Between the East and West?
In China, the eastern regions (such as Guangdong) have high demand for computing power but face challenges such as high land costs and limited access to renewable energy. The western regions (such as Inner Mongolia), on the other hand, have abundant renewable energy but insufficient computing power demand. Computing-power synergy addresses this issue through the "East Data, West Computing" initiative, which allows data from the east to be processed in the west, utilizing the region's renewable energy resources. This approach achieves a win-win situation by meeting the eastern regions' computing needs while reducing excess energy in the west.
5. How Is the Government Responding to the Growth in Electricity Consumption for Computing Power?
The government has clearly defined the direction for computing-power synergy as "strengthening computing power with electricity and promoting electricity development through computing power." On one hand, it is coordinating the development of energy resources and computing facilities to ensure a stable power supply. On the other hand, it is encouraging the participation of computing power in power grid management, transforming data centers from mere electricity consumers into resources that can help manage demand fluctuations. In the future, measures such as cross-regional power exchanges and virtual power plants will be implemented to balance the growth of computing power with the stability of the power system.
Overall, the increase in electricity consumption for computing power is an inevitable outcome of digital development, and computing-power synergy is key to balancing this growth with the supply of electricity. It not only supports the development of AI and data industries but also drives the transformation of the power system towards greater efficiency and sustainability.