A September 8 report from BloombergNEF forecasts that China’s operational data center IT capacity will reach 45.5 GW by 2030, more than doubling 2025 levels, with over half of projects under construction or in planning already located in northern and northwestern regions. Hubs such as Ulanqab are rapidly aggregating AI compute, but building capacity is not the same as using it well. As the center of gravity in computing supply shifts west, cross-regional scheduling and matching compute with...

BloombergNEF’s September 8 report points to a clear trend: by 2030, China’s operational data center IT capacity will reach 45.5 GW, more than doubling from 2025. More notable is the structural shift—more than half of projects currently under construction and planned are in the north and northwest, with computing hubs in Inner Mongolia, Ningxia, and Gansu becoming the main engines of expansion.

For years, China’s data center market was defined by “oversupply in the west and unmet demand in the east.” As the “East Data, West Computing” initiative advances, eight national hubs position Beijing-Tianjin-Hebei, the Yangtze River Delta, the Greater Bay Area, and Chengdu-Chongqing as demand hubs, while four western regions serve as supply hubs. Policy is also explicitly guiding large data centers toward areas with more abundant land and power.

The west’s appeal is tangible: lower electricity prices, ample land, and fiber networks connecting to major demand centers. Based on media reports and institutional research, the combined cost of operating data centers in Gansu, Ningxia, and Inner Mongolia can be 30%