We mapped every data centre on earth

We mapped every data centre on earth

by International Energy Agency

Applied Sciences & Engineering Economics & Finance KNTMedia, entertainment, information and communication industriesKNTXInformation technology industries
🎙 International Energy Agency 👥 28K 📅 August 4, 2026 ⏱ 15 min 👁 34K 📄 documentary 🧭 2026-08-16
Available in: English (current) Français

Keywords

data centresAIelectricityenergygrid

Summary

The video, produced by the International Energy Agency, investigates the rapid expansion of AI data centres and their impact on global energy systems. It highlights that in 2025, tech companies invested over $400 billion in data centre infrastructure, with projections of $700 billion in 2026. Using satellite tracking, the IEA mapped data centre buildout, finding that installed capacity doubled in 12 months. The video explains that a single server rack can now consume up to 600 kW, equivalent to 65 households, and hyperscale data centres use as much energy as 100,000 homes. Globally, data centres could consume nearly 1,000 TWh by 2030, about 3% of world demand. The analysis shows data centres cluster near cities, straining local grids, and can use up to 30% of local electricity supply. The video discusses challenges such as rapid power demand spikes, grid connection delays, and the need for on-site generation and storage. It also notes that AI efficiency is improving, but agentic AI could offset gains. Finally, it explores AI’s potential to help the energy sector, estimating savings of 13.5 exajoules by 2035, and distinguishes between energy-intensive general-purpose AI and efficient, purpose-built models.

190 words

Critical Evaluation

Value of the Information & Strength of the Argument

The video provides valuable, data-driven insights into the energy implications of AI data centres, based on original geospatial analysis by the IEA. It presents a balanced view, acknowledging both challenges and opportunities. The argumentation is solid, supported by specific examples (e.g., 400% power spike in a quarter second, largest battery project) and clear explanations of complex concepts like grid capacity and price dynamics. The use of satellite data adds a novel, empirical dimension. The video effectively communicates the scale of investment and its potential consequences, while also highlighting AI’s potential to improve energy systems. The reasoning is coherent and well-structured, moving from investment to infrastructure, to grid impacts, to solutions and future outlook.

Scientific Rigor, Source Quality, Title Accuracy

The video demonstrates high scientific rigor, drawing on the IEA’s latest analysis and original geospatial research. It references specific data points and provides links to the IEA’s Energy and AI observatory for further details. The sources are credible and authoritative. The title accurately reflects the content, as the video indeed maps data centres globally. The video also includes expert commentary from IEA analysts, enhancing its credibility. The presentation is professional and objective, with no apparent bias. The inclusion of chapters helps structure the content logically. Overall, the video maintains high standards of accuracy and reliability.

223 words

Title / Content Match

The title accurately reflects the content, as the video presents a global mapping of data centres based on satellite tracking and geospatial analysis.

Quality & Reliability

9/10

The video is produced by the International Energy Agency (IEA), a reputable intergovernmental organization, and is based on their latest analysis and original geospatial research. It presents data with clear methodology and references to their Energy and AI observatory, enhancing credibility.

Chapters

Cited Sources

Concurring Sources

  • IEA - Energy and AI — The video is based on IEA analysis, and this page provides additional reports and data.

Contribution & Novelties

The video provides a unique, up-to-date global mapping of AI data centres using satellite tracking, which is a novel contribution to understanding the scale and distribution of this infrastructure. It offers original analysis on clustering patterns and their implications for local grids. The video also contextualizes data centre energy use within broader electricity demand growth, providing a balanced perspective. It highlights emerging solutions like on-site generation and storage, and discusses the potential of AI to improve energy efficiency. The distinction between general-purpose and purpose-built AI models is an important insight for policy and technology discussions.

Pour aller plus loin :

  • International Energy Agency - Energy and AI — Official IEA topic page with reports and data.
  • Data centre energy consumption — Overview of data centre energy use on Wikipedia.
  • Small modular reactors — Background on SMRs, a potential power source for data centres.

143 words

Radar Profile

The radar profile shows high scores in information quantity, quality, and reliability, with a slightly lower technical level, indicating that the video is highly informative and trustworthy, but accessible to a general audience. The balance between technical depth and clarity is well managed.

Reliability 9/10

💬 The comments are predominantly positive, with viewers expressing appreciation for the informative content and impressive satellite data. Some comments raise concerns about the environmental impact and cost of data centres, while others ask for more detailed maps or data. Overall, the sentiment is engaged and constructive, with a mix of praise and critical questions.