Summary of Key Points
With the explosion of large AI models and the computing power industry, China's consumption of electricity for computing purposes is growing at a rate of 100 billion kilowatt-hours per year, and it is expected to account for 6% of the total national electricity usage by 2030. However, there is a mismatch between the concentrated demand for computing power in the eastern regions and the scarcity of green energy there, while the western regions have abundant green energy but lack the necessary computing power applications. The traditional model of "power grids supplying electricity and computing power consuming it" is becoming unsustainable. "Coordinated management of electricity and computing power" (using digital scheduling to dynamically match the two) has become a critical solution. There are already cases in Guangdong where data centers are participating in electricity trading through virtual power plants, but challenges remain, such as the rigid or flexible nature of energy consumption, unclear market mechanisms, and barriers between different departments. In the long run, this represents a systematic transformation of the electricity market, computing power industry, and energy structure.
Detailed Analysis
Why is Coordinated Management of Electricity and Computing Power an Urgent Issue?
The computing power industry is a major consumer of electricity: The cost of electricity accounts for 40%-60% of the operating expenses for data centers, and the stability of power supply and electricity prices directly affect their profitability. For example, eastern cities have a high concentration of computing power centers but face shortages of land and green energy resources; western regions, on the other hand, generate plenty of renewable energy but lack the necessary computing capacity. Take Shaoguan as an example—a node for "eastern data processing with western computing power." It attracts computing companies due to its surplus of electricity and the lowest industrial electricity prices in the province. Without coordinated management, the high costs of electricity in the east would make it unviable for these centers to operate, while the green energy in the west would be wasted, resulting in losses for both regions.
What Exactly is Coordinated Management of Electricity and Computing Power?
In simple terms, it involves "matching electricity supply with computing demand dynamically." This can be achieved through digital means by adjusting the allocation of green energy from the west with the computing needs in the east, as well as varying electricity prices and computing tasks across different times. For instance, during periods of high wind and solar output in the west, non-urgent computing tasks (such as video rendering or data backup) can be moved to the west, and data centers in the east can reduce their power consumption during peak usage times. This approach addresses three main issues: where to obtain electricity (giving priority to green energy from the west), how to ensure the stability of the power grid (with the help of storage and scheduling), and how to manage computing tasks (distinguishing between essential and non-essential operations).
How Can We Transform Energy Consumers into Flexible Resources?
Data centers, once passive consumers of electricity, can now become more strategic entities:
- Participation in Electricity Trading through Virtual Power Plants: Data centers owned by Guangdong Unicom and Mobile are participating in spot electricity trading via virtual power plant platforms, allowing them to use more electricity when prices are low and reduce non-essential operations when prices are high, marking the first time computing capacity has been adjusted in response to market changes.
- Integration of Energy Generation, Grid, Consumption, Storage, and Computing: For example, the Gui'an data center uses storage systems to balance energy supply and demand; it stores excess green energy during periods of surplus and releases it when needed. It also employs direct current (DC) technology to ensure uninterrupted power supply even under extreme conditions, with local renewable energy meeting over 80% of its electricity needs.
- Cross-regional Scheduling: Through advanced systems, non-real-time computing tasks in Guangzhou can be moved to areas with abundant green energy in Guizhou within minutes, saving on electricity costs and supporting the consumption of western renewable energy.
What Are the Current Challenges to Coordinated Management of Electricity and Computing Power?
The process is still in its early stages, facing three major hurdles:
- Technical Barriers: It is difficult to distinguish between essential and non-essential operations that can be adjusted. Core services (like real-time communications and security) must not be interrupted, while non-real-time tasks can be more easily reconfigured.
- Market Risks: Participating in spot electricity trading involves price fluctuations and potential penalties for deviations from planned usage, which deter data centers from engaging deeply.
- Interdepartmental Cooperation: Electricity is regulated by the Energy Bureau, while computing power is managed by the National Data Administration. The lack of coordination between these departments creates significant barriers to effective collaboration.
What Long-term Benefits Can Coordinated Management of Electricity and Computing Power Bring?
This approach is not just about optimizing electricity usage; it represents a revolution in three key areas:
- More Efficient Electricity Markets: By allowing new types of consumers (such as data centers) to participate in trading, price signals become more accurate, leading to better resource allocation.
- Smarter Computing Power Industries: Computing power can be used more strategically, making electricity costs more manageable and generating additional revenue through power management.
- A Greener Energy Structure: Computing power consumption can be matched more precisely with renewable energy generation, maximizing the use of new energy sources and contributing to the achievement of carbon neutrality goals.
In summary, coordinated management of electricity and computing power is a critical intersection for digital economy and energy transformation. It aims to ensure the sustainability of the computing power industry while minimizing waste of green energy, ultimately achieving a win-win situation with stronger computing capabilities, greener energy use, and lower costs. However, this goal requires significant technological advancements, improved mechanisms, and enhanced interdepartmental cooperation—a challenging but essential path to follow.