Chinese Grid Operators Question Feasibility of Renewables for AI Data Centers

Chinese grid operators are expressing skepticism about the government's ambitious plans to increase renewable energy usage in powering artificial intelligence (AI) data centers. According to industry analysts and officials cited by Reuters, the push to have renewables supply a majority of electricity demand at these facilities by 2030 faces significant hurdles. The primary concern is the unpredictability of peak power demands from AI data centers, which are critical for operating high-performance computing devices like GPUs.

Data center operators often aim to maximize usage of expensive GPU hardware as quickly and intensively as possible, complicating efforts to match renewable energy supply with demand patterns. This challenge is further exacerbated by the fact that China's current electricity mix for data centers is heavily reliant on coal, accounting for nearly 70% of the total power generation as of 2025, according to a report from the International Energy Agency (IEA). The IEA also projects that solar and wind energy will contribute significantly to the grid by 2030, but only if provincial mandates and policies promoting data center construction in renewable-rich western regions are effectively implemented.

Despite these challenges, China has already taken steps towards integrating renewables into its digital infrastructure. Recently, the country inaugurated an innovative underwater data center off the coast of Shanghai that utilizes offshore wind power for cooling and electricity generation. This facility aims to reduce environmental impacts by minimizing energy consumption, water usage, and land requirements. However, achieving broader integration of renewable sources remains a complex task due to technical and operational constraints within the data center sector.