Chinese green technology company Envision commissioned what it claims is the world's largest single AI data center building — a 120,000-square-meter (1,291,668-square-foot, roughly 29.7-acre) supercomputing facility in Ulanqab, Inner Mongolia. The campus is built around a novel architecture: instead of connecting to an overloaded power grid, it wires renewable generation assets directly to the compute facility, sidestepping the interconnection queues that are stalling Western AI infrastructure by five years or more.

That distinction matters more than the building's size. According to Goldman Sachs Research, the United States faces a structural data center power shortfall of approximately 9.3 gigawatts in 2026, widening to an estimated 45 gigawatts by 2028 — and grid interconnection queues in Northern Virginia, Phoenix, and Dallas are now running four to seven years. The primary reason is not money or chips. It is electricity — specifically, the wait for permission to draw it from the grid. Envision's Ulanqab model proposes a different path: go where the energy already is, build the generation assets next to the servers, and skip the queue entirely.

How the AI Power System Works

The technical foundation of Galaxy Campus is Envision's proprietary AI Power System, a platform that coordinates renewable generation, battery energy storage, dedicated transmission infrastructure, and computing consumption within a single operating system.

Most hyperscaler data centers in Western markets claim "100% renewable energy" through virtual power purchase agreements — contracts in which clean electricity is generated somewhere else and matched to the facility's consumption on paper. The Galaxy Campus model is architecturally different: generation assets are physically co-located with the compute infrastructure, wired directly rather than routed through a regional grid. Independent technical research confirms that this approach matters. DC-coupled architectures — which connect generators and storage directly to computing loads without AC conversion stages — reduce transmission losses and enable what researchers call "islanded operation," allowing the facility to maintain stable power delivery even when regional grid conditions fluctuate. Vertiv chief product and technology officer Scott Armul has described DC power as inherently better suited to renewables than traditional AC grid connections.

The AI Power System coordinates all of this. Wind and solar generation output flows to battery storage systems, which smooth intermittency and maintain the stable power delivery that AI training workloads require. Ricky Zheng, General Manager of Envision's AI Data Center division, said the integrated architecture enables the facility to deliver "up to ten times more compute output per square meter than conventional data centers" — a claim attributed to Envision and not independently audited. "Supporting a million accelerators requires a purpose-built AI Power System that integrates renewable generation, dedicated transmission infrastructure, and large-scale energy storage to deliver stable, low-cost green power," Zheng said, according to Envision's official press release.

Why Inner Mongolia, and Why Now

Ulanqab sits approximately 350 kilometers (217 miles) northwest of Beijing. Its selection as the campus site is strategic in ways that extend well beyond land cost.

The city's average annual temperature is approximately 4°C (39°F), which substantially reduces mechanical cooling requirements — one of the largest operational costs in any data center. Electricity is priced at around 0.32 yuan per kilowatt-hour (approximately $0.05 USD/kWh) — roughly one-third to one-fifth of industrial electricity prices in US markets under grid-connection pressure. The energy mix already draws more than 80% from renewables: Inner Mongolia installed 135 gigawatts of combined wind and solar capacity in 2024, surpassing the region's coal power capacity for the first time and leading China in total, new, and generated renewable capacity. Despite the distance from Beijing, network latency remains below five milliseconds — sufficient for demanding AI inference workloads.

Ulanqab is not a blank slate for data infrastructure. It is one of eight national computing hubs under China's "Eastern Data, Western Computing" initiative, the central government's program to shift compute-intensive workloads from resource-constrained eastern cities to infrastructure-rich western regions. The Horinger Data Center Cluster — which includes major parks in Hohhot and Ulanqab — carried 291,000 petaflops of total computing capacity and 273,000 petaflops of intelligent computing capacity at last official count, with green electricity usage running at 84.57% — a national benchmark.

Envision is now the most ambitious operator to plant its flag in that framework. DeepSeek, the Chinese AI startup, separately announced plans this week for its own gigawatt-scale campus in Ulanqab, underscoring that the region has crossed from regional ambition to genuine hyperscaler destination.

The Infrastructure Race Galaxy Campus Enters

The commissioning arrives at a moment when scale — not willingness to spend — is the binding constraint on AI development. According to Epoch AI research, five data centers at gigawatt scale or above are expected to come online globally in 2026, each operated by a different hyperscaler. The largest currently operational is xAI's Colossus 2 in Memphis, drawing approximately 946 megawatts.

The spending behind this buildout is staggering. Amazon, Google, Meta, and Microsoft combined for $413 billion in capital expenditures in 2025 — an 84% increase from $224 billion in 2024 — and are projected to spend $600 billion to $700 billion in 2026. McKinsey has estimated that companies across the compute-power value chain will need to invest $5.2 trillion into data centers by 2030 to meet worldwide AI demand, according to McKinsey's compute cost analysis.

Yet capital has not solved the power access problem. Grid interconnection queues across the US hold more than 2,600 gigawatts of pending projects, with average wait times of five years in many regions and four to seven years in the most data-center-dense markets. Up to 30-50% of planned 2026 US data center capacity is expected to slip to 2028 because of those constraints. Every major Western hyperscaler has responded by signing nuclear or small modular reactor power deals to secure supply outside the standard interconnection queue.

Envision's approach is structurally different from both the nuclear bets and the conventional grid-dependent model. By co-locating renewable generation with compute in a region already rich in wind and solar resources, Galaxy Campus bypasses the queue problem at its root.

Mission Gobi: Beyond a Single Building

Galaxy Campus is the first flagship project under Envision's Mission Gobi initiative, which the company unveiled at the VivaTech conference in Paris in June 2026. The initiative targets 5 gigawatts of green AI computing capacity in desert and arid regions worldwide by 2030.

The ambition reflects Envision's position in the global energy industry. The company ranked second globally by wind turbine installations in 2025, shipping 20.9 gigawatts — nearly a quarter installed outside mainland China — according to BloombergNEF's 2025 market report. In July 2026, S&P Global named Envision to its Tier 1 Cleantech Companies list for the second consecutive year, with Tier 1 recognition across wind turbines, energy storage systems, and battery cells. The company has also appeared twice on Fortune's "Change the World" list and was named to TIME's 100 Most Influential Companies in 2024.

The Galaxy Campus pivot — using renewable energy expertise not just to generate clean power for others, but to directly run some of the world's most power-hungry compute workloads — represents a significant expansion of Envision's market position.

What Operating Under Chinese Law Means for Data on This Infrastructure

Envision Energy is headquartered in Shanghai and, as a company operating under Chinese jurisdiction, is subject to China's national security legislation. Readers and enterprise buyers evaluating this infrastructure should understand the legal framework before making deployment decisions.

China's National Intelligence Law (2017), Article 7, requires that "all organizations and citizens shall support, assist, and cooperate with national intelligence efforts in accordance with law." This obligation applies to Chinese companies — including Envision Energy — regardless of where their servers are physically located or what their stated privacy policies say. The law contains no exceptions for private companies and no mechanism by which a company can legally refuse an intelligence request and publicly disclose that refusal. The US Department of Homeland Security has explicitly warned that this law empowers the Chinese government to require PRC entities to provide data from foreign persons or businesses, or to compel companies to install backdoors in equipment.

China's Cybersecurity Law (2016) Article 28, amended in January 2026, further requires network operators to provide "technical support and assistance" to public security and national security agencies on demand. The Data Security Law (2021) adds cross-border data transfer restrictions and additional government access provisions.

These are fixed legal conditions of operating under Chinese jurisdiction, not contested claims. Envision has not been the subject of US government security findings or CFIUS actions of the kind that have affected Huawei, and some legal scholars argue that the National Intelligence Law's practical application is narrower than commonly characterized. But enterprise buyers placing workloads on Galaxy Campus or future Mission Gobi facilities should treat the legal framework as a structural condition of that decision, independent of Envision's contractual privacy commitments or the location of specific servers.

Capacity on Paper vs. Capacity in Operation

The commissioning marks Galaxy Campus's first operational phase. The headline figures — 2 gigawatts of capacity, one million AI accelerators, one million PFLOPS — represent the campus's designed ceiling, not its current output. Envision has not disclosed the capacity at which the facility will initially operate, and "commissioning" in the context of a phased infrastructure build typically refers to the first operational phase becoming live.

That gap between announced capacity and delivered capacity is worth watching. In the broader Western data center market, 30-50% of planned 2026 capacity is expected to slip to 2028. Galaxy Campus avoids the grid queue problem, but it faces its own scaling challenge: populating one million accelerator slots requires access to AI chips, and China's domestic chip ecosystem — dependent on Huawei Ascend series chips built on SMIC's 7nm process — delivers less throughput per chip than Western alternatives.

Whether the renewable energy architecture, the available compute, and the declared capacity can converge in a single functional building at 2GW scale is a question that will be answered over the next several years of buildout.

Exchange rate used: 1 USD = 6.749 CNY (mid-market, August 6, 2026). All currency conversions are approximate.

Frequently Asked Questions

What makes Galaxy Campus different from other data centers claiming 100% renewable energy?

Most hyperscaler data centers achieve "100% renewable energy" status through virtual power purchase agreements, which match clean electricity generated at a separate location to the facility's consumption on paper. Galaxy Campus uses direct physical co-location: renewable generation assets are wired to the facility's computing infrastructure rather than routed through a regional grid. This distinction matters operationally — physical co-location eliminates grid interconnection dependency and transmission losses, while virtual matching leaves the facility still drawing from whatever power the local grid provides.

Why is the power grid the main bottleneck for AI data centers in the US?

Grid interconnection queues across the United States hold more than 2,600 gigawatts of pending projects, with average wait times of five years in many regions and four to seven years in the highest-density data center markets like Northern Virginia and Phoenix. Transformer delivery times have stretched to three to five years, and switchgear is sold out through 2028. Goldman Sachs Research estimates a structural US power shortfall of 9.3 gigawatts in 2026, widening to 45 gigawatts by 2028. The result: 30-50% of planned 2026 US data center capacity is expected to slip to 2028. Money does not solve this problem — grid physics and infrastructure lead times do.

Does data stored or processed at Galaxy Campus fall under Chinese government access laws?

Yes, as a structural legal condition. China's National Intelligence Law (2017), Article 7, requires all Chinese organizations to cooperate with national intelligence work on demand. China's Cybersecurity Law (Article 28) requires network operators to provide technical assistance to public security agencies on demand. These obligations apply to Envision Energy as a Chinese company regardless of contractual privacy provisions, server location, or Envision's stated policies. The US Department of Homeland Security has specifically warned that PRC laws give the Chinese government authority to demand data from PRC-operated data centers. Enterprise buyers considering deploying sensitive workloads on this infrastructure should treat the legal framework as a fixed condition of that decision.

What is China's Eastern Data, Western Computing initiative, and why does it matter here?

Launched in 2022 by China's National Development and Reform Commission, the "Eastern Data, Western Computing" initiative redirects compute-intensive workloads from energy-constrained eastern cities to eight designated computing hubs in western and northern China, including Inner Mongolia. The initiative provides policy support, preferential electricity pricing, and infrastructure investment to regions with abundant renewable energy and available land. Inner Mongolia is one of the eight hubs, and Ulanqab specifically has attracted dozens of data center projects under this framework. Galaxy Campus is the most ambitious single facility built within this initiative to date, but it is not alone: DeepSeek and Chindata Group have each separately announced major compute builds in the same region.