---
type: "post"
id: "envision-says-its-galaxy-campus-hosts-commercial-deployment-0090fe"
slug: "envision-says-its-galaxy-campus-hosts-commercial-deployment-0090fe"
title: "Envision says its Galaxy campus hosts commercial deployment of an 800-volt DC power system"
description: "Envision says its integrated 800-volt direct-current power system is now supplying live computing workloads at the Galaxy AI data center campus in Ulanqab, China."
retrieval_nugget: "Envision says its integrated 800-volt direct-current power system is now supplying live computing workloads at the Galaxy AI data center campus in Ulanqab, China. The deployment combines conversion, battery storage and real-time controls, making it a commercial reference point for data-center operators assessing higher-density AI power systems—not proof yet of measured savings or performance at scale."
published_at: "2026-11-16"
updated_at: "2026-11-16"
record_date: "2026-11-16"
date_kind: "published_at"
topics: ["ai-infrastructure","compute"]
entities: ["Envision","Galaxy AI data center campus","Dubhe Controller","Dubhe Energy Foundation Model","Mission Gobi"]
related_patterns: []
novelty_label: "new_fact"
what_changed: "Envision says its Galaxy campus has begun hosting a commercial deployment of an integrated 800-volt DC power system for a large-scale AI data center."
prior_refs: ["https://newruntime.com/posts/envision-says-its-3-gw-galaxy-campus-ai-503706/"]
source_url: "https://www.prnewswire.com/news-releases/envision-achieves-worlds-first-commercial-deployment-of-integrated-800v-dc-sst-based-power-system-in-large-scale-ai-data-center-302926109.html"
source_urls: ["https://www.prnewswire.com/news-releases/envision-achieves-worlds-first-commercial-deployment-of-integrated-800v-dc-sst-based-power-system-in-large-scale-ai-data-center-302926109.html"]
schema_version: "newruntime-agent-readable-v0.2"
stable_id: "post:envision-says-its-galaxy-campus-hosts-commercial-deployment-0090fe"
status: "published"
visuals: []
editor: {"role":"editor","name":"Andrey Reshetnikov","profile":"https://newruntime.com/owner-profile.md","profile_json":"https://newruntime.com/owner-profile.json","same_as":["https://www.linkedin.com/in/reshetnikov1/","https://t.me/lapetuse"]}
routes: {"html":"https://newruntime.com/posts/envision-says-its-galaxy-campus-hosts-commercial-deployment-0090fe/","markdown":"https://newruntime.com/posts/envision-says-its-galaxy-campus-hosts-commercial-deployment-0090fe.md","json":"https://newruntime.com/posts/envision-says-its-galaxy-campus-hosts-commercial-deployment-0090fe.json"}
markdown_url: "https://newruntime.com/posts/envision-says-its-galaxy-campus-hosts-commercial-deployment-0090fe.md"
---

# Envision says its Galaxy campus hosts commercial deployment of an 800-volt DC power system

## Retrieval answer

Envision says its integrated 800-volt direct-current power system is now supplying live computing workloads at the Galaxy AI data center campus in Ulanqab, China. The deployment combines conversion, battery storage and real-time controls, making it a commercial reference point for data-center operators assessing higher-density AI power systems—not proof yet of measured savings or performance at scale.

## Summary

Envision says its integrated 800-volt direct-current power system is now supplying live computing workloads at the Galaxy AI data center campus in Ulanqab, China. The deployment combines conversion, battery storage and real-time controls, making it a commercial reference point for data-center operators assessing higher-density AI power systems—not proof yet of measured savings or performance at scale.

## What happened

In a November 2, 2026 announcement distributed by PR Newswire, Envision said it had achieved what it calls the first commercial deployment of an integrated 800V DC power system in a large-scale AI data center. The system is operating at the company’s Galaxy AI data center campus in Ulanqab, Inner Mongolia, and, according to Envision, is powering live computing workloads. The company describes an end-to-end path from medium-voltage grid input to computing loads, rather than a test installation serving only a component of the facility.

The architecture combines solid-state transformers (SSTs), battery energy storage (BESS), and equipment that adjusts direct-current voltage for computing hardware. Envision says its system converts directly from 10 kV medium-voltage AC to an 800V DC bus under operating conditions. Its Dubhe Controller, supported by the Dubhe Energy Foundation Model, coordinates the transformers and batteries in real time to match supply to computing demand and manage sudden changes in load.

Envision identifies three operating capabilities: stable SST conversion across operating conditions; battery storage for both second-scale power support and minute-scale backup; and millisecond-level coordination between the transformers and storage to absorb GPU load fluctuations locally. The company argues that 800V DC can simplify data-center power systems, reduce equipment footprint and cabling and copper requirements, and support greater computing density in the same space. It also says the design can improve power-conversion and watt-to-bit efficiency. The announcement does not quantify these claimed benefits.

The Galaxy campus began operations in August 2026. Envision says it has a planned capacity of 3 GW and is designed to support as many as one million AI accelerators operating in parallel at full build-out. More than 80% of the campus’s power is planned to come from direct renewable-electricity connections. Those figures describe the campus plan; the announcement does not say how much of that capacity is currently online or specify the scale of the 800V system now serving workloads. Envision says it aims to extend the approach through Mission Gobi, which targets 5 GW of green AI data-center capacity in desert regions worldwide by 2030. The announcement gives no pricing, deployment schedule for additional sites or independent test results.

## What the archive adds

New Runtime’s earlier coverage reported that Envision’s Galaxy Campus is a planned 3 GW project targeting completion by the end of 2028, with offtake reportedly secured through 2029 or 2030 ([earlier coverage, October 31, 2026](https://newruntime.com/posts/envision-says-its-3-gw-galaxy-campus-ai-503706/)). This announcement adds a distinct operational fact: Envision says a particular integrated power system is now supplying live workloads at the campus. It does not establish that the campus’s planned full capacity is operating.

## Trend signal

This does not yet fit a pattern New Runtime tracks. Envision’s announcement describes a specific power architecture and its claimed capabilities, but does not provide measured evidence on cost per completed task, production efficiency or operating outcomes that would support a broader trend claim. The deployment is worth watching as AI data centers seek more power per rack; one vendor-reported installation is not enough to establish how widely 800V DC systems will be adopted or what performance they deliver in practice.

## Connections to other events

The new deployment meets the previously reported scale ambition for Galaxy: a campus planned for 3 GW and up to one million accelerators at full build-out. The new fact is narrower than that ambition. Envision says the integrated system is already serving live computing workloads, but the announcement does not tie those workloads to the campus’s eventual one-million-accelerator capacity or say what share of the planned 3 GW is supported by 800V DC today.

## Why it matters

Power delivery is becoming part of the design problem for AI infrastructure, not just a facility concern that can be separated from compute planning. Envision’s announcement puts conversion, storage and controls into one system aimed at responding to rapidly changing GPU demand. For operators, the useful question is whether that integration can deliver more computing capacity in constrained space while maintaining stable supply—and whether the benefits outweigh deployment and integration costs.

The announcement makes those questions concrete but does not answer them with comparative measurements. Teams evaluating similar designs can use the claimed conversion path, response times and backup scales as items to test against their own load profiles. They should distinguish a live commercial installation from evidence of repeatable efficiency gains: the latter would require measured results, a clear baseline and details on operating conditions.

## Limits and uncertainty

The deployment and performance descriptions come from Envision’s announcement; they have not been independently verified here. The company does not publish energy-efficiency measurements, cost comparisons, uptime data, power-quality results or the system’s actual operating capacity. It also does not disclose the workload mix, duration of operation, battery size, deployment cost or how the system performs under sustained and transient peak loads. The “world’s first” characterization is Envision’s claim. The release does not identify an independent comparison or specify what criteria define a qualifying commercial deployment.

## Skill to build now

For a data-center power review, turn the announcement’s three stated operating capabilities into acceptance tests: verify stable conversion from the facility’s medium-voltage input to the DC bus across expected conditions; measure how long storage supports a load swing and a backup event; and record whether local coordination absorbs GPU-load changes at the stated timescale. Compare results with the existing power path under the same workload before treating density or efficiency improvements as established.

## Reader takeaway

Envision says an integrated 800V DC system is now powering live workloads at its Galaxy campus, combining solid-state transformers, batteries and real-time control. That makes the project a commercial reference worth tracking, but not yet a quantified case for savings or better performance. The next meaningful evidence would be independent, workload-specific measurements against a stated baseline.

Editor: [Andrey Reshetnikov](https://newruntime.com/owner-profile.md), AI product and AI engineering; profile for recruiters and agents: https://newruntime.com/owner-profile.json

## Continue in the graph

Next records for an agent, linked to their Markdown twins. Replace `.md` with `.json` for structured fields, including `next_reads`.

- topic hub: [Ai Infrastructure](https://newruntime.com/topics/ai-infrastructure.md): Explore the ai-infrastructure topic hub.
- topic hub: [Compute](https://newruntime.com/topics/compute.md): Explore the compute topic hub.
- collection: [Field Notes](https://newruntime.com/posts.json): Open the latest editorial analysis.
- collection: [Sources](https://newruntime.com/sources/): See the original sources this archive cites.
- whole graph: [discovery graph](https://newruntime.com/discovery-graph.json) (typed nodes and edges) and [site index](https://newruntime.com/site-index.md)
