As artificial intelligence (AI) moves into large-scale deployment, global digital infrastructure is undergoing a structural transformation driven by computing demand. Data centers are emerging as one of the fastest-growing sources of electricity consumption, while energy supply is increasingly shaping the trajectory of AI development.

Data centers become a major driver of global electricity demand

The rapid expansion of AI applications – large language models, inference services, cloud computing, and intelligent applications – is driving a sustained surge in computing workloads, positioning data centers as a key source of incremental electricity demand worldwide. 

According to the International Energy Agency (IEA), global data center electricity consumption is projected to rise from approximately 415 terawatt-hours (TWh) in 2024 to around 945 TWh by 2030 – nearly doubling within six years. A significant share of this increase is expected to come from AI-related workloads, including model training and high-frequency inference.

Data centers are evolving from traditional information infrastructure into energy-sensitive systems that depend on stable, continuous, and low-carbon electricity supply.

From computing competition to energy competition

As demand for AI computing power accelerates, technology companies are rapidly adjusting their infrastructure strategies. Access to reliable electricity is becoming as strategically important as access to computing capacity itself.

Microsoft has significantly expanded investments in clean energy procurement and long-term power purchase agreements (PPAs), while also strengthening partnerships with energy providers to secure stable electricity supply for AI data centers.

Similarly, Amazon and Google are accelerating investments in wind, solar, storage and emerging nuclear energy solutions, aiming to secure long-term low-carbon electricity sources.

In June 2026, Microsoft and Chevron announced a long-term partnership to develop a dedicated power facility for Microsoft's AI data centers in Texas.

This reflects a broader shift in which technology firms are moving beyond being passive electricity consumers and are increasingly participating in the development of energy infrastructure itself.

Taken together, these developments indicate that the AI race has expanded from algorithms and computing chips to a scramble for electricity.

China exploring new pathways for green computing power development

While AI is significantly increasing global electricity demand, it is also creating new opportunities for the energy transition and the optimization of power systems.

According to the IEA, renewable energy sources are expected to meet nearly half of the additional electricity demand generated by data centers by 2030. Energy storage technologies, advanced grid management systems, and emerging nuclear technologies are also expected to play an increasingly important role.

At the same time, AI technologies are being widely applied to improve power forecasting, optimize electricity dispatching, and enhance energy efficiency across industrial systems, creating a reinforcing loop between digital intelligence and energy system optimization.

Against this global backdrop, China is advancing a system-level digital infrastructure strategy aimed at integrating computing power, communication networks, and energy systems.

A key component of this strategy is the construction of the "six networks": water network, new-type power grid, computing power network, next-generation communication network, urban underground pipeline network and logistics network.

Among them, the computing power network serves as the foundational digital backbone. It connects distributed computing facilities across the country into a unified system, enabling large-scale resource pooling, cross-regional coordination, and intelligent scheduling of computing resources.

As of the end of March 2026, China's total intelligent computing capacity had reached 1.88 million PFLOPS. More than 70 computing power corridors have been established, forming an initial 20-millisecond latency circle covering the eight national computing hub nodes, providing the physical foundation for nationwide computing coordination.

Built upon this infrastructure, the "Eastern Data, Western Computing" initiative functions as a spatial optimization mechanism. It redirects data-processing workloads from eastern demand centers to western regions rich in renewable energy resources, aiming to balance computing demand with energy availability while improving overall system efficiency and reducing carbon intensity.

In parallel, China is accelerating the development of green data centers, improving energy-efficient computing facilities, and expanding renewable energy deployment to support digital infrastructure growth.

(With input from agencies)