344
Productivity & Workflow355
Automation & Workflow224
Software Development251
Marketing & Growth192
AI Infrastructure & MLOps174
Writing & Content Creation203
Data & Analytics141
Design & Creative170
Photography & Imaging156
Customer Support131
Sales & Outreach125
Voice & Speech135
Education & Learning131
Operations & Admin87
Envision founder Zhang Lei says AI data centers should move off the power grid and run in deserts on dedicated wind, solar, and batteries.
In short: A Chinese clean energy company is testing AI data centers in desert areas that run on their own wind and solar power instead of the main electricity grid.
AI data centers, meaning large buildings full of computers that run AI services, use a huge amount of electricity. In many places, that power comes from the public grid, the same system that supplies homes and businesses. This has led to worries about higher power prices, strained local grids, and higher carbon emissions.
Zhang Lei, founder of Envision, says the fix is to build more AI data centers in deserts and power them with dedicated “microgrids” (a small, separate power system). He argues these sites should use wind and solar power, plus large batteries to store energy so the computers can run day and night. Think of it like a factory building its own private power station, instead of plugging into the city.
Envision says it has built what it calls the world’s first AI data center powered entirely by off-grid renewable energy at a site in Inner Mongolia, developed with Tencent. The company has not disclosed the facility’s size. Envision is also building a larger “gigawatt-scale” data center in the same region and Zhang says Envision plans to build 5 gigawatts of off-grid renewable powered data centers by 2030.
Two big questions are water and reliability. Data centers often use water for cooling, and deserts are dry, so Envision says it will rely on “dry” and closed-loop cooling that uses much less water. The other issue is whether batteries and extra renewable generation can keep power steady at the near-constant level large customers expect.
Source: Financial Times