AI Updates
  • Latest News
  • Models & Research
    • AGI & Alignment
    • AI Agents
    • Foundation Models
    • Generative AI
    • Open Source AI
    • Research & Papers
    • Rankings & Benchmarks
  • Infrastructure
    • Chips & Semiconductors
    • Cloud & Networking
    • Data Centers & Neoclouds
    • Developer Platforms
    • Quantum Computing
  • Business
    • Big Tech Strategy
    • Mergers & Acquisitions
    • Public Markets
    • Startups & Funding
    • Talent & Leadership
    • Market Analysis
  • Applications
    • Consumer & Creative
    • Cybersecurity
    • Education & Media
    • Enterprise & Productivity
    • Finance & Legal
    • Healthcare & Life Sciences
    • Industry & Manufacturing
    • Science & Discovery
  • Robotics
    • Autonomous Vehicles
    • Drones & Defense
    • Edge AI & Embodied Intelligence
    • Robotics
  • Energy
    • Cooling & Water
    • Efficiency & Emissions
    • Electricity & Power Grids
    • Nuclear & Renewables
  • Policy & Society
    • AI Safety & Security
    • Copyright & Misinformation
    • Geopolitics
    • Privacy & Human Rights
    • Regulation & Government
    • Work & Society
Stay Ahead of AI

Get the AI developments that matter, delivered to your inbox.

Follow AI Updates
Facebook
Twitter
Instagram
  • Home
  • About
  • News
  • Contact
  • Newsletter
0
0
0
AI Updates AI Updates
AI Updates
  • Latest News
  • Models & Research
    • AGI & Alignment
    • AI Agents
    • Foundation Models
    • Generative AI
    • Open Source AI
    • Research & Papers
    • Rankings & Benchmarks
  • Infrastructure
    • Chips & Semiconductors
    • Cloud & Networking
    • Data Centers & Neoclouds
    • Developer Platforms
    • Quantum Computing
  • Business
    • Big Tech Strategy
    • Mergers & Acquisitions
    • Public Markets
    • Startups & Funding
    • Talent & Leadership
    • Market Analysis
  • Applications
    • Consumer & Creative
    • Cybersecurity
    • Education & Media
    • Enterprise & Productivity
    • Finance & Legal
    • Healthcare & Life Sciences
    • Industry & Manufacturing
    • Science & Discovery
  • Robotics
    • Autonomous Vehicles
    • Drones & Defense
    • Edge AI & Embodied Intelligence
    • Robotics
  • Energy
    • Cooling & Water
    • Efficiency & Emissions
    • Electricity & Power Grids
    • Nuclear & Renewables
  • Policy & Society
    • AI Safety & Security
    • Copyright & Misinformation
    • Geopolitics
    • Privacy & Human Rights
    • Regulation & Government
    • Work & Society
Large data centres on a cold Inner Mongolian plateau beside wind power and dry terrain
  • Data Centers & Neoclouds

Ulanqab Emerges as China’s AI Data Centre Hub Despite Water Pressure

  • August 22, 2026
  • No comments
Total
0
Shares
0
0
0
0

Ulanqab, a high-altitude city in Inner Mongolia, is becoming one of Asia’s fastest-growing centres for AI computing. Nearly 100 data centres have opened or entered construction there since 2016, while Chinese companies have announced projects representing an estimated 12.5 gigawatts of combined capacity.

More than 70% of those commitments were announced during the past year, according to a Goldman Sachs research note cited by WIRED. The scale signals a change in how Chinese AI developers obtain computing power: companies that historically rented capacity are beginning to invest directly in large physical infrastructure.

Cold weather and cheap power draw investment

Ulanqab’s location offers several advantages. Long, cold winters reduce the energy needed for cooling, and electricity in Inner Mongolia is among the least expensive in China. The region has expanded wind and solar generation while retaining abundant coal supplies.

The city is also close enough to Beijing for many workloads. Two dedicated fibre links completed in 2017 and 2019 reduced average latency to under five milliseconds, according to WIRED. That performance can support real-time services such as inference, while model training is generally less sensitive to delay because a run can continue for months without constant user interaction.

Chinese technology companies including DeepSeek, ByteDance, Alibaba and Xiaohongshu are reported to be building facilities in the area. DeepSeek’s project is described as particularly large, although WIRED does not provide a confirmed capacity or completion date for it.

Inner Mongolia has hosted data centres for a decade. Huawei opened one in Ulanqab in 2016 and Apple followed in 2019. The region later became part of the national “Eastern Data, Western Compute” strategy, which directs computing infrastructure toward China’s western interior. AI training and growing domestic demand have now given remote capacity a more commercially valuable role.

Water scarcity complicates the buildout

The same geography that provides cold air and inexpensive land also creates an environmental constraint. Ulanqab receives roughly 14 inches of rain annually, and the local government is already struggling to meet residents’ water demand. Last month, the municipal water company shut several waterworks for seven hours each night to reduce peak pressure.

Data centres in the city reportedly need additional cooling water during only two months of the year. That seasonal advantage may reduce annual consumption compared with warmer locations, but dozens of planned facilities could still add significant demand during the driest period.

WIRED does not provide a combined water forecast for the announced projects, so their effect on household supply cannot yet be quantified from the article. Commitments may also be delayed, redesigned or never completed. The current restrictions nevertheless show that infrastructure planning is already operating against a limited resource.

The issue complicates any simple claim that Inner Mongolia is an ideal low-cost location. Electricity and cooling efficiency matter, but a compute cluster also depends on reliable water, transmission lines, fibre and local consent. Capacity measured in gigawatts does not reveal whether all those supporting systems are ready.

Renewable growth still shares the grid with coal

Chinese officials see data centres as a way to consume renewable electricity that might otherwise be curtailed. Envision, a major wind-turbine manufacturer, announced a two-gigawatt Ulanqab AI facility this month that it says will connect directly to its own clean-power supply.

Researchers cited by WIRED have found a relationship between regions with unused renewable capacity and regions building data centres. Continuous computing demand could improve the economics of wind and solar projects by providing a large customer near the source of generation.

The electricity mix is not yet clean. Research cited in the article estimates that coal still supplies about 37% of Ulanqab’s power. Data centres operate around the clock, which has historically encouraged reliance on dispatchable fossil generation when wind or sunlight is unavailable.

Experts say the balance could shift quickly as Inner Mongolia adds renewables, but a fully renewable future remains a projection. Ulanqab’s boom demonstrates both China’s accelerating compute investment and the physical trade-offs beneath it: low-cost energy, cold weather and fibre connectivity on one side, water scarcity and continuing coal dependence on the other.

Post Views: 98
Total
0
Shares
Share 0
Tweet 0
Pin it 0
Share 0
Topics
  • ChinaAI
  • DataCentres
  • Energy
  • Ulanqab
Previous Article
An orbital AI computing satellite above Earth with constrained launch routes in the distance
  • Data Centers & Neoclouds

Starcloud Adds $250 Million as Orbital Data Centres Face a Launch Squeeze

  • August 21, 2026
View Post
Next Article
Rows of AI processors in a data center connected to a structured financial bridge
  • Data Centers & Neoclouds

Lambda Adds $1 Billion in Debt to Expand Its AI Chip Fleet

  • August 29, 2026
View Post

Stay Ahead of AI

Get the AI developments that matter, delivered to your inbox.

More AI News
Editorial illustration of a creative studio desk with abstract image panels and a film strip emerging from a simple light source, representing a startup pivot into AI-generated campaign assets
View Post
  • 3 min
  • Startups & Funding

Melius Raises a $20 Million Series A After Rebuilding Its Advertising AI Business

  • October 7, 2026
Conceptual photograph of a clean data-center aisle with server cabinets and a concentrated warm light at the far end, representing infrastructure growth and concentrated demand
View Post
  • 3 min
  • Data Centers & Neoclouds

Lambda Seeks Up to $4 Billion as Its IPO Plans Meet a Concentrated Backlog

  • October 7, 2026
Minimal editorial illustration of a small luminous courier-like form approaching a translucent gate, representing permission-based access for an AI agent
View Post
  • 3 min
  • AI Agents

Website Permissions Are Becoming a Bottleneck for Consumer AI Agents

  • October 7, 2026
Illustrative photograph of a researcher examining abstract geometric research sheets at a sunlit desk, representing careful review of AI-assisted mathematics
View Post
  • 3 min
  • Research & Papers

OpenAI’s Planned Math Release Reopens a Dispute Over Scientific Disclosure

  • October 7, 2026
Abstract editorial illustration of an open geometric lattice and flowing translucent forms, representing a model preview and future access to weights
View Post
  • 3 min
  • Foundation Models

Mistral Large 4 Preview Separates Benchmark Claims From Its Planned Open-Weight Release

  • October 6, 2026
Conceptual illustration of a small connected infrastructure cluster inside layered translucent boundaries, representing data control rather than proven security
View Post
  • 3 min
  • AI Safety & Security

Malaysia Links Its AI Nation Ambition to Data Security and Digital Trust at CyberDSA

  • October 6, 2026
Illustrative photograph of a network technician inspecting fibre connections in an unbranded telecommunications equipment room
View Post
  • 3 min
  • Enterprise & Productivity

Deutsche Telekom Sets AI Savings Targets While Planning Reinvestment and Revenue Growth

  • October 6, 2026
Conceptual editorial photograph of paper-like layers revealing a subtle granular pattern under angled light, not an actual text watermark or detector
View Post
  • 3 min
  • Copyright & Misinformation

OpenAI’s EU Text Watermark Plan Highlights Limits of AI Authorship Detection

  • October 6, 2026

Leave a Reply Cancel reply

Your email address will not be published. Required fields are marked *

Subscribe

Subscribe now to our newsletter

AI Updates
  • About
  • Newsletter
  • Contact
  • Privacy Policy
  • Terms of Use
Artificial Intelligence News

Input your search keywords and press Enter.

Go to mobile version