China's data centres will consume 774 terawatt-hours of electricity by 2030, quadrupling current usage as artificial intelligence workloads trigger a structural jump in computing-related power requirements, according to a new report from Wood Mackenzie titled "Rewiring China's grid for the AI era." By 2060, these facilities could represent 17% of the nation's total electricity consumption, up from just 2% today. The analysis highlights how AI's rapid expansion is forcing China to rethink the relationship between computing infrastructure and the power grid.
Data centre demand will triple its share of total power consumption to 6% by 2030, the report notes, and continue climbing to 17% by 2060. Eight national computing hubs are expected to control over 70% of total data centre capacity through 2060, though grid limitations and renewable energy growth may gradually push some expansion to power-rich regions outside these designated zones. Carbon emissions from data centres are forecast to peak by 2035 as green power mandates tighten and renewable penetration increases. AI training is driving more energy-intensive infrastructure, while the rapid spread of inference is creating a growing baseline of continuous electricity demand that will keep data centre power consumption rising even as China's broader electricity demand growth slows.
"Data centres are becoming an increasingly important part of China's energy system," said Wanting Zhao, research analyst for Asia Pacific power and renewables 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 report finds that China's 15th Five-Year Plan calls for coordinated development of computing infrastructure and green electricity, pushing policies that promote direct green power supply and more flexible computing loads. According to Wood Mackenzie, new power architectures, battery storage, and intelligent workload scheduling could transform data centres into demand-side flexibility providers rather than passive base loads.
The report explains that China is shifting from a "power follows compute" model to "compute follows power," particularly for workloads that can tolerate latency. Historically, data centres clustered near major cities to access customers, skilled workers, network infrastructure, and low-latency connectivity. But AI's expansion has introduced a new priority: the availability of abundant, reliable, and low-carbon electricity. Renewable-rich regions in western China offer plentiful land and growing renewable generation capacity, creating opportunities to place power-intensive AI training, batch processing, and data storage closer to electricity sources. Eastern demand centres will likely retain an important role for latency-sensitive applications such as AI inference and financial services, creating a division of labour between eastern computing demand hubs and western power-rich computing centres. For workloads that don't require immediate processing, computing demand can shift to time windows with abundant renewable generation or lower electricity prices, granting computing loads significant temporal flexibility and allowing data centres to absorb curtailed renewables in oversupplied regions.
The report concludes that data centre demand for green electricity will rise substantially through 2060, but warns that unlocking compute-power synergy requires addressing data security and service-level agreement concerns during cross-regional migration, stimulating operator and tenant participation, and implementing targeted mechanisms to ease heavy initial capital expenditure burdens. "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," Zhao concluded. The next challenge extends far beyond physical grid planning into the commercial and regulatory frameworks that will determine whether China can successfully align its computing infrastructure with its clean energy goals.