Free tools Windows power users keep installed
One-click scans. No signup required.
Hyperscale data centres are still expanding because demand—especially for AI computing—and investment remain strong. But investment and announced projects are not the same as completed capacity: grid connections, power equipment, chips, permits, financing and local acceptance can delay or reshape what gets built. Land is a potential siting issue, but the available global evidence does not measure its independent effect.
What does “growth” mean in this context?
Several different measures are often described as data-centre growth: companies’ capital spending, electricity consumed by operating facilities, and proposed projects. Each shows a different part of the story. Spending signals commitment, electricity use reflects facilities already running, and proposals indicate development ambition—not delivered capacity.
| Measure | Reported figure | How to interpret it |
|---|---|---|
| Capital expenditure | The International Energy Agency (IEA) says five large technology companies spent more than USD 400 billion in 2025 and estimates their spending will rise another 75% in 2026. IEA, 2026 | This is investment by five companies, not a count of new data centres or a measure of completed capacity. The 2026 figure is an estimate. |
| Electricity use | Data-centre electricity consumption increased 17% in 2025; consumption at AI-focused centres increased 50%. The IEA reports both changes in its 2026 outlook. IEA, 2026 | These figures describe electricity consumption, not the number of facilities built during the year. |
| Projected electricity use | The IEA’s central outlook puts data-centre consumption at 485 TWh in 2025 and projects 950 TWh in 2030—around 3% of global electricity demand by 2030. IEA, 2026 | The 2030 value is a projection, not a realised outcome. |
| Large-project proposals | In the first half of 2026, North America represented 64% and EMEA 21% of identified proposed power for new data centres and campuses above 100 MW. Uptime Institute, 2026 | This is a proposal dataset for very large projects, not a census of completed hyperscale facilities. Proposed power may not be built on the announced schedule—or at all. |
Why does demand remain strong?
AI is increasing demand for data-centre capacity, both to train models and to run them for users. The IEA reports that data-centre electricity consumption grew in 2025, with the increase at AI-focused centres substantially faster. Uptime Institute’s 2026 operator survey also describes strong demand, increasingly driven by high-density and AI-related workloads. IEA outlook; Uptime Institute survey, published July 24, 2026
Efficiency gains do not automatically cancel out this growth. The IEA says efficiency per task is improving, while energy-intensive uses are expanding. If the number or intensity of computing tasks rises faster than the energy needed for each task falls, total electricity use can still increase. The IEA’s outlook reflects that net effect rather than assuming that efficiency alone will determine demand. IEA outlook
Do these 3 things before closing this tab:
1Repair Windows errors before they cause bigger problems2Fix the driver behind crashes, sound loss and screen glitches3Clear out junk files and repair common Windows errors#1 Best Overall
What can prevent proposed capacity from becoming operational?
Electricity is a central constraint, but a connection to the grid is only one part of project delivery. The IEA identifies bottlenecks involving grid connections, energy technology supply chains, advanced chips, regulatory capacity, financing and community acceptance. Uptime Institute’s operator survey adds concerns about power availability and reliability, rising costs, supply chains and staffing. These pressures can overlap: a project might have a site and financing yet still wait for a connection, equipment or skilled workers. IEA outlook; Uptime Institute survey
- Grid access and reliability: Connection timing and available transmission capacity can limit when a site can draw the power it needs. The IEA notes that most data centres prefer connecting to the grid, even as some developers consider onsite generation. IEA outlook
- Power equipment and generation: Transforming, transmitting and supplying large amounts of reliable power depends on equipment and supply chains. Onsite gas generation is not a guaranteed quick workaround: the IEA says some U.S. developers are considering it, but turbine supply is constrained and reliable service for critical, variable loads can require overbuilding generation.
- Chips and other technology: A building and power supply are not enough if advanced chips or other essential components are unavailable. The IEA identifies chip manufacturing and energy technology supply chains among the constraints on expansion. IEA outlook
- Permitting and local acceptance: Regulatory capacity and community acceptance affect whether projects can proceed. They are distinct from physical land availability: a site may exist while approvals or local support remain unresolved. IEA outlook
- Operations and workforce: Uptime Institute respondents report staffing shortages as well as power, cost and supply-chain challenges. The survey also reports continuing cooling constraints and rising peak rack densities, which can complicate operation as computing loads intensify. These are survey findings, not a forecast of how much capacity will be built. Uptime Institute survey
The IEA says onsite gas generation may be overbuilt by 30%–70% relative to demand to provide reliable power for critical, variable data-centre loads. This is a reported range for that reliability approach, not a universal requirement for every facility. It illustrates why adding onsite generation does not necessarily eliminate the need to plan carefully for capacity and reliability. IEA outlook
Rank #2
- Ideal for Gifting
- Ideal for a bookworm
- Compact for travelling
How significant is land scarcity?
Large campuses plainly need suitable sites, and siting involves more than acreage: approvals and community acceptance can also matter. However, the IEA outlook and Uptime Institute materials cited here do not provide a comparable global measure of available land, land costs, or the number of projects delayed specifically because land could not be secured. They therefore do not establish land scarcity as the dominant global bottleneck or allow it to be ranked against grid access.
For a particular project, land should be assessed alongside connection timing, transmission availability, firm power, permitting, cooling and water resources where documented, equipment lead times, staffing and financing. The available sources do not provide a unified country-by-country scorecard for these factors. IEA outlook; Uptime Institute proposal analysis
How should regional growth figures be read?
Uptime Institute’s proposal analysis shows the geographic distribution of identified proposed power for very large projects, not the distribution of operating capacity. The IEA also publishes a chart of global data-centre capacity additions and capacity at risk of connection delay over 2025–2030, but its page does not state the chart’s numerical values in the accompanying text. No at-risk total should be inferred from that text alone. Uptime Institute proposal analysis; IEA grid-constraint chart
For any region or individual project, the useful question is not simply how much capacity has been announced. Separate proposals from facilities in operation, then examine when a grid connection and firm power are available, whether permits and community acceptance are in place, and whether equipment, chips, workers and financing can be secured. A project’s exposure to these constraints depends on its location and design, so a global proposal pipeline cannot by itself show how much capacity will arrive, or when.
Quick Recap
Best Value
- It can be a gift option
- Comes with secure packaging
- Helpful in various ways
Rank #4
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
