The continued growth in AI computing demand is changing the energy requirements of data centres. Intelligent computing clusters require not only large-scale and stable electricity supplies, but also place new demands on power system planning, construction, and operational management in terms of load fluctuations, energy costs, and green electricity consumption.

Against this backdrop, power-computing coordination is gradually moving from a development concept toward practical engineering applications. By strengthening coordination between computing infrastructure and power systems and integrating energy supply, load management, intelligent dispatching, and green electricity trading, this approach can help improve resource allocation efficiency and support the continued development of the AI industry.

1. Moving from Power Supply Assurance to Two-Way Coordination Between Supply and Demand

Traditional data centre development has primarily focused on electricity connection capacity, supply reliability, and backup power configuration. As the scale of AI training and inference tasks expands, simply increasing power supply capacity is no longer sufficient to address the computing industry’s combined requirements for energy efficiency, green electricity, and operating costs.

Power-computing coordination emphasises systematic optimisation across multiple areas, including electricity supply, computing scheduling, and energy management.

On the power side, grid facilities and power supply solutions need to be planned rationally based on regional grid capacity, the distribution of clean energy resources, and projected electricity demand growth. On the computing side, more flexible approaches to task scheduling and resource allocation can be explored according to task priorities, time requirements, and the geographical distribution of computing resources.

When computing workloads have a degree of temporal and geographical flexibility, some tasks can be rescheduled or relocated in accordance with electricity prices, green electricity availability, and grid operating conditions, provided that service quality, data security, and business timeliness requirements are met.

This means that computing is no longer merely a load that passively receives electricity. Under appropriate conditions, it may also become a resource capable of participating in power system regulation.

2. Intelligent Dispatching Is an Important Technological Foundation for Power-Computing Coordination

Power-computing coordination requires comprehensive analysis of electricity supply and demand, computing loads, renewable energy output, and equipment operating conditions. Digital monitoring, AI algorithms, communication networks, and energy management platforms can provide technological support for this process.

In practice, the main applications of these technologies include the following areas.

• Electricity Load Forecasting and Operational Monitoring

By collecting information on electricity loads, equipment operating conditions, and renewable energy output, forecasting models can be used to analyse changes in electricity demand and provide a basis for grid planning, power supply assurance, and load management.

• Coordinating Computing Tasks with Energy Supply

Subject to business requirements, the execution time or deployment location of certain computing tasks can be optimised according to factors such as electricity supply conditions, green electricity availability, and market prices. This can improve the alignment between computing resources and energy resources.

• Intelligent Power System Management

Digital technologies can support grid condition monitoring, equipment performance analysis, fault warnings, and dispatching decisions, improving the observability and operational management efficiency of power supply systems.

• Energy Management and Efficiency Optimisation

By taking into account the operating conditions of data centre cooling systems, power supply and distribution systems, and computing equipment, operators can monitor energy consumption, analyse efficiency, and optimise operating strategies to reduce unnecessary energy use.

It should be noted that computing dispatching cannot replace grid safety constraints or professional power system dispatching. Relevant systems must comply with electricity operating procedures, cybersecurity requirements, and computing service quality standards to achieve safe and reliable coordinated operation.

3. Direct Green Power Supply and Source-Grid-Load-Storage Coordination Expand Application Opportunities

The utilisation of clean energy is one of the key directions of power-computing coordination.

Wind and solar power offer low-carbon advantages, but their output is affected by weather, time, and regional resource conditions. If computing facilities can be appropriately coordinated with renewable energy projects, the power grid, and energy storage systems, there may be opportunities to improve the utilisation of green electricity.

Models such as direct green power supply, green electricity trading, and integrated source-grid-load-storage coordination provide different pathways for exploring these opportunities.

Direct green power supply focuses on establishing a more direct connection between renewable energy generation and electricity demand through power supply facilities and operating mechanisms that comply with applicable requirements. Green electricity trading, by contrast, enables electricity consumers to purchase qualifying green electricity through market-based mechanisms. Source-grid-load-storage coordination takes a broader approach by considering the coordinated operation of power sources, grids, loads, and energy storage.

In actual projects, the appropriate model must be designed in accordance with grid connection conditions, project investment requirements, energy supply reliability, trading rules, and applicable policies.

In suitable applications, energy storage can help address temporal mismatches between renewable energy output and electricity demand, supporting power supply assurance, load regulation, and renewable energy utilisation. However, the required storage capacity and economic viability must still be evaluated based on actual project conditions.

4. Reliable Power Supply and Power Quality Remain Fundamental Requirements

For high-density computing equipment, a stable electricity supply is essential to maintaining continuous data centre operations. The concentrated growth of computing loads may increase pressure on local power grids, making electricity connection capacity, power supply architecture, and equipment operation and management increasingly important.

Therefore, alongside the advancement of power-computing coordination, fundamental work remains necessary in areas such as grid expansion and upgrades, substation construction, distribution system optimisation, protection and control, and backup power supply planning.

For individual projects, the characteristics of the equipment and the system design must be taken into account when assessing power quality, harmonics, reactive power, short-circuit current, protection coordination, and supply reliability. Equipment selection and configuration should be based on engineering design, technical standards, and actual operating conditions.

Digital dispatching and power equipment development are complementary rather than interchangeable. Together, they form the foundation of power-computing coordination: digital technologies help systems monitor and coordinate resources more effectively, while physical power infrastructure provides the basis for stable electricity supply and safe operation.

5. Integrating Technological Innovation with Engineering Practice

Power-computing coordination is a systematic undertaking involving energy, electricity, communications, computing, and digital management. Its development requires not only innovation in intelligent algorithms and dispatching platforms, but also continued upgrades to grid infrastructure, energy management technologies, and related equipment.

In the future, as computing facilities expand, renewable energy utilisation improves, and electricity market mechanisms develop further, power-computing coordination is expected to play a greater role in green energy supply, load management, and system operational optimisation.

DKNC will continue to follow developments in new-type power systems, electrical equipment technologies, and energy digitalisation. The company will also monitor the supporting role of power infrastructure in the development of the computing industry, as well as the evolution of related technologies and application scenarios.

Through technological innovation, engineering practice, and industrial value-chain collaboration, power-computing coordination can help improve the allocation of electricity and computing resources, supporting the green, reliable, and sustainable development of digital infrastructure.

Sources: National Energy Administration, Xinhua News Agency, People’s Daily, Digital China Summit