Electricity consumption by China’s data centers is expected to nearly quadruple by 2030, driven primarily by the rapid expansion of artificial intelligence (AI) computing, according to a recent industry report. The report projects that data centers will use approximately 774 terawatt-hours of electricity by the end of the decade, representing about 6 percent of the country’s total power consumption, a threefold increase from current levels.

The surge reflects the increasing demand for AI training, which requires high power density and substantial upgrades to existing power infrastructure. Additionally, the widespread adoption of AI inference applications has created a consistent, round-the-clock load on the power grid. This growth in data center power usage is notable given that overall national electricity demand is moderating.

Data from China’s National Energy Administration shows that electricity consumption for internet data services increased by 41 percent year-on-year, reaching 49.4 billion kilowatt-hours in the first half of the year. Wu Liqiang, an official with the China Electricity Council, noted that electricity demand from emerging sectors continued to grow rapidly during the first six months of 2026 and is expected to maintain this trajectory. Power use in China’s high-tech and equipment manufacturing sector rose by 9.8 percent year-on-year to 600.8 billion kilowatt-hours during the same period, accounting for nearly a quarter of manufacturing electricity consumption.

The report highlights the importance of access to reliable, affordable, and low-carbon electricity for locating new data centers. China’s eight national computing hubs are projected to retain their dominance, hosting over 70 percent of total computing capacity through 2060. However, capacity limits in these core areas mean some new facilities will be developed in regions rich in renewable energy, particularly in China’s western provinces, which offer abundant land and expanding clean energy resources. Meanwhile, eastern economic centers are expected to continue managing latency-sensitive applications, including real-time AI processing and financial services.

Policy support is playing a critical role in facilitating this transition. The country’s 15th Five-Year Plan (2026-30) emphasizes the coordinated development of computing infrastructure and green energy. Measures such as direct green power supply incentives and cross-regional resource scheduling are intended to bolster the integration of renewable energy with data center operations.

The evolving energy landscape is also enabling data centers to become more flexible, shifting away from being “rigid, passive loads” toward functioning as adaptive resources within the power system. This transformation involves combining new power supply models, battery energy storage, and intelligent workload management to enhance operational elasticity.

As China increases its share of renewable energy and strengthens green power mandates, the report forecasts that carbon emissions from the data center sector will peak by 2035. Despite rising demand for green electricity well into 2060, this shift aims to balance growth with sustainability objectives.