China’s data centers are on track to quadruple their electricity consumption, reaching 774 terawatt-hours (TWh) by 2030 as artificial intelligence workloads drive a fundamental surge in computing power demand.
By 2060, facility energy consumption could account for as much as 17% of the nation’s total electricity usage, according to new research from Wood Mackenzie titled Rewiring China’s grid for the AI era.
“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.”
The surge in energy consumption is being propelled by a combination of accelerating AI demands and aggressive national policy targets:
- AI Workload Bifurcation: Resource-heavy AI model training requires high-intensity energy inputs, while widespread AI inference deployment creates a expanding, continuous baseline of electricity usage.
- National Policy Directives: China’s 15th Five-Year Plan explicitly demands “compute-power synergy”—the synchronized expansion of computing infrastructure with green electricity—supported by mandates for direct renewable power supply and flexible load deployment.
This shifting power dynamic is reshaping the physical footprint of Chinese computing infrastructure. Historically, facilities were built near high-density urban centers to minimize latency and access local talent and infrastructure. To meet expanding AI energy demands, the industry is pivoting from a “power follows compute” model to a “compute follows power” strategy for workloads where latency is not a factor.
Under this new framework, latency-tolerant operations—such as model training, data storage, and batch processing—are shifting toward land- and resource-rich western provinces. Latency-sensitive applications, including real-time AI inference and financial transaction processing, will remain clustered near primary eastern demand centers.
While eight designated national hubs are set to maintain dominant positions in capacity through 2060, regional grid limitations and expanding renewable power generation are expected to spur ongoing development in power-rich areas outside these core zones.
To manage this growing load, data centers are evolving into flexible, demand-side grid management resources through several technological implementations:
- Integration of modern power architectures and utility-scale battery storage.
- Implementation of intelligent workload scheduling to shift flexible processing to times of peak renewable generation or lower wholesale power prices.
- Functioning as dynamic power sinks to absorb excess curtailed generation in oversupplied renewable regions.
Wood Mackenzie projects that data center reliance on green electricity will expand significantly through 2060. Consequently, carbon emissions from China’s data centers are expected to peak by 2035 as higher volumes of renewable energy enter the national grid mix.
“The next challenge extends far beyond physical grid planning,” Zhao concluded. “To truly unlock compute-power synergy, the industry must address data security and SLA concerns during cross-regional migration, stimulate the willingness of operators and tenants to actively participate, and implement targeted mechanisms to alleviate the heavy initial CAPEX burden. The direction of travel is clear: computing is beginning to follow power, but our commercial, security, and institutional architectures must urgently evolve to make this commercially viable.”
You can find more information about Wood Mackenzie’s data centre market analysis here: Data Centre Market Analysis & Trends | Wood Mackenzie
Leave a Reply