China data centre power demand to quadruple to 774 TWh by 2030: Report

China's data centre power demand to quadruple to 774 TWh by 2030 amid AI boom

New Delhi, August 25

The research firm said data centres could account for 6 per cent of China's total electricity consumption by 2030, up from current levels, with their share potentially rising to 17 per cent by 2060.

"Data centres are becoming an increasingly important part of China's energy system," said Wanting Zhao, research analyst, Asia Pacific power and renewables research at Wood Mackenzie.

"As AI drives sustained growth in computing demand, access to reliable, cost-competitive and lower-carbon electricity will play a growing role in determining where and how new data-centre capacity is developed," Zhao said.

Wood Mackenzie said AI training is driving demand for more energy-intensive infrastructure, while the rapid expansion of AI inference is adding a growing base of sustained electricity demand. As a result, data-centre power consumption is expected to continue rising even as broader electricity demand growth in China moderates.

The release said this is also changing the traditional relationship between computing infrastructure and power availability. Data centres have historically clustered around major cities and technology hubs to access customers, talent and network infrastructure.

However, the growth of AI is increasingly encouraging a "compute follows power" model, particularly for workloads that can tolerate higher latency.

Renewable-rich regions in western China could attract power-intensive AI training, batch processing and data storage, while eastern centres are likely to retain an important role in latency-sensitive applications such as AI inference and financial services.

Wood Mackenzie expects eight national hubs to remain the dominant centres of China's computing capacity through 2060, although grid constraints and rising renewable generation could encourage gradual expansion into other renewable-rich regions.

The firm also expects data centres to play a more active role in the power system. Battery storage and intelligent workload scheduling could allow non-urgent computing tasks to be shifted to periods of abundant renewable power or lower electricity prices.

Data-centre demand for green electricity is expected to rise substantially through 2060, while their carbon emissions are forecast to peak by 2035 as renewable power use increases.

However, Zhao said wider adoption of the compute-power model would require changes beyond physical grid planning, including addressing data security, service-level concerns, operator participation and the high initial capital costs involved.

— ANI