Berliner Tageszeitung - Envision Deploys World’s First Commercial 800V DC Power System for AI Data Centers

China’s Envision Energy has announced the commercial operation of an integrated 800-volt direct-current (DC) power architecture at an artificial intelligence facility in Inner Mongolia, marking what the company describes as a world-first deployment of this configuration.

The installation is located at the Galaxy AI campus in Ulanqab, where trials of its Dubhe controller and supporting energy model began in August 2026. According to Envision, the system represents a significant shift from conventional data center power designs, which typically convert incoming electricity through multiple stages before delivering power suitable for servers and accelerators.

Envision’s design employs a solid-state transformer to convert 10-kilovolt medium-voltage alternating current directly into an 800-volt DC distribution line. The company states that this approach reduces the number of conversion steps, thereby improving electrical efficiency and allowing power equipment to occupy substantially less physical space within data centers.

A key advantage cited by Envision is the potential for reduced material usage. A higher-voltage distribution network requires less copper for cabling, a factor the company says could lower costs as computing installations expand rapidly. The architecture also incorporates battery storage designed to respond rapidly to processor demand changes and provide temporary backup during unexpected interruptions. Envision claims the system can provide immediate support over seconds and maintain backup capability for several minutes.

The deployment includes a controller that coordinates the transformer and batteries within milliseconds, aiming to manage sudden fluctuations in electricity consumption from graphics processing units (GPUs). Such coordination is critical because AI workloads can cause substantial swings in demand, placing additional pressure on electrical infrastructure supplying densely packed computing equipment. Envision also notes that the system can accommodate higher-density server racks, allowing operators to install more computing capacity within a given floor area.

The Galaxy facility is expected to grow to 3 gigawatts of capacity, with Envision stating its eventual scale could serve one million AI accelerators. The company intends to introduce the technology through Mission Gobi, an initiative seeking 5 gigawatts of renewable-powered computing capacity by 2030. This program focuses on desert locations, where Envision expects renewable generation and energy storage to support large computing developments.

The system has been designed to work alongside renewable electricity sources, with its batteries intended to balance supply and changing computing requirements. However, Envision has not provided independently verified measurements quantifying efficiency improvements, installation costs, or savings compared with conventional power systems. The announcement also does not establish how the architecture performs under sustained full-load conditions or across facilities with substantially different hardware configurations.

While the deployment marks a commercial milestone for Envision, the scale of its claimed operational advantages awaits detailed independent evidence.

S. Soerensen--BTZ