China Builds AI Map of its Entire Wind and Solar Grid

China Builds AI Map of its Entire Wind and Solar Grid

A new wave of concern is spreading across major economies as artificial intelligence systems rapidly increase global electricity demand, putting pressure on power grids that were never designed for today’s compute-heavy infrastructure.

In the United States, electricity capacity prices in the PJM Interconnection, the largest grid operator in the country, have surged more than tenfold over the past two years. Analysts point to the rapid expansion of large-scale data centers as a key factor driving the spike. In Europe, utilities are facing similar strain, with grid operators accelerating transmission upgrades to keep up with demand from hyperscale computing facilities.

The International Energy Agency (IEA) warns that global electricity consumption from data centers could reach nearly 1,000 terawatt-hours by 2030, fueled largely by artificial intelligence workloads. While renewable energy capacity is expanding, many countries still lack the systems needed to efficiently coordinate variable sources like wind and solar at national scale using AI-driven grid optimization.

Against this backdrop, researchers in China have unveiled a major development that could reshape how renewable energy systems are managed.

A study published in the journal Nature, led by researchers from Peking University and Alibaba Group DAMO Academy, presents what is described as the first complete, high-resolution AI-generated national inventory of wind and solar infrastructure. The system maps nearly the entire renewable energy landscape of China using deep learning applied to satellite imagery.

Using a model trained on sub-meter resolution satellite data, the team identified 319,972 solar photovoltaic installations and 91,609 wind turbines across the country. The process analyzed more than 7.5 terabytes of imagery, producing a unified dataset that spans 1,915 counties.

Researchers say this level of visibility allows renewable energy to be studied and managed as a single interconnected system rather than fragmented provincial grids. One of the key findings is that solar and wind power tend to complement each other more effectively when coordinated over larger geographic areas. For example, weather conditions that reduce solar output in one region may not affect wind generation in another, allowing overall supply to remain more stable when integrated nationally.

The study suggests that China’s current provincial-level grid coordination limits the full benefits of this complementarity. A shift toward national-scale optimization could reduce wasted energy, known as curtailment, which remains a significant issue in renewable-heavy regions.

Professor Liu Yu of Peking University described the system as giving grid operators a “complete overview” of the country’s renewable energy assets, enabling a level of planning that was previously impossible.

This development comes as China experiences a sharp rise in electricity demand from artificial intelligence and digital infrastructure. According to the China Electricity Council, power consumption from data centers and related services rose by 44 percent year-on-year in early 2026, reaching 22.9 billion kilowatt-hours.

Much of the new data center construction is concentrated in northern and western provinces, where land is cheaper and wind and solar resources are more abundant. These regions also show the strongest potential for renewable complementarity, making them strategic hubs for both clean energy and AI infrastructure growth.

Beyond its immediate findings, the study highlights the broader potential of large-scale geospatial AI. By turning raw satellite data into structured national energy maps, researchers demonstrated a framework that could be replicated in other countries seeking to better integrate renewable energy systems.

China’s clean energy sector, already massive in scale, generated an estimated 15.4 trillion yuan (about 2.26 trillion US dollars) in economic output last year, according to the Finland-based Centre for Research on Energy and Clean Air. As artificial intelligence continues to expand global electricity demand, the ability to map, coordinate, and optimize renewable infrastructure may become a defining factor in future energy security.