The company’s latest offerings focus on solving the reliability bottleneck currently facing AI developers. Central to this is the AI Power System, which utilizes physical AI models to synchronize generation, storage, and computing loads. By incorporating 2.5 MW solid-state transformer technology, the system converts standard grid inputs into 800V DC power with 98.5% efficiency, reducing copper requirements by 80%. To handle the volatile load profiles of modern GPU clusters, Envision pairs lithium iron phosphate storage for multi-hour shifting with sodium-ion batteries for rapid, millisecond-level power smoothing.
Beyond data centers, Envision introduced its Gen 8 Long-Duration Energy Storage System. Engineered for grid-forming applications, the unit supports up to 16 hours of duration and maintains 250% overload capability to stabilize renewable-heavy grids. The hardware is built to withstand extreme temperatures from -40°C to 50°C, offering a 25-year operational lifespan. Kevin Huang, Senior Vice President at Envision, emphasized that the next phase of the energy transition relies on the competitiveness of the entire power system rather than just cost per watt-hour. These technologies are already seeing field application, notably at the Chifeng Net-Zero Industrial Park and the forthcoming gigawatt-scale Envision Galaxy Campus in Ulanqab.

Comments (0)
No comments yet. Be the first!