Epoch AI · Sep 16, 2026
Epoch AI expands its AI data center research
Markdown · JSON · Original source
Epoch AI’s post published September 16, 2026, describes an expanded AI Data Centers explorer and a redesign for viewing facility data. The organization says its research covers an estimated 44% of global AI compute across 86 data centers, and includes satellite imagery and details such as compute capacity, chip types, capital costs, and construction status. The explorer lets users compare sites over time and filter or recolor charts by owner, primary user, or country. Epoch AI says its coverage estimate compares compute capacity at operational data centers with estimated compute delivered to major customers, assuming a three-month deployment lag. It also says the data captures most of Meta’s AI compute, while much of Microsoft’s and Amazon’s compute is not yet included. Epoch AI identifies public-cloud deployments as difficult to attribute and quantify because infrastructure may serve several customers. It reports relatively low coverage of Chinese data centers and says it aims to improve that coverage. The announcement describes a growing dataset and new ways to inspect individual sites; it does not establish how complete the estimates are across companies or regions.
Headline and summary by Escape Meridian, prepared by Luna from the attributed source. A source check is not independent confirmation of the source’s claims.
Original source
Epoch AI · Article
Source details
Original source date: Sep 16, 2026 (2026-09-16T00:00:00.000Z).
Date basis: Date supplied by the publisher’s feed.
Added to Escape Meridian: 2026-10-29T16:58:17.174Z. First observed: 2026-10-18T02:47:41.814Z. Retrieved: 2026-10-23T21:15:38.106Z.
Referenced by
Recorded citations and public file attachments. A reference alone does not establish agreement or influence.
No public references recorded here yet.
Limits
Text as retrieved on October 23, 2026; originally published September 16, 2026. The coverage figures are estimates based on Epoch AI’s stated method; the post notes gaps in company and regional coverage and difficulty attributing public-cloud infrastructure. The post describes the explorer and its data coverage but does not independently establish how complete its estimates are.
Open question
How will Epoch AI’s coverage estimates change as it captures more public-cloud and Chinese data center capacity?
Source-stated
Epoch AI says the explorer covers an estimated 44% of global AI compute across 86 data centers.
Today, our research covers an estimated 44% of global AI compute, spread over 86 data centers.
Source-stated
Epoch AI says facility records include satellite imagery and data on compute capacity, chip types, capital costs, and construction status.
For each site, we provide satellite imagery and detailed facility data, including compute capacity in H100-equivalents, chip type(s), capital costs, and construction status.
Source-stated
Epoch AI estimates coverage by comparing operational data center compute capacity with estimated compute delivered to major customers, assuming a three-month deployment lag.
We estimate coverage by adding up all the compute capacity of operational data centers (in H100-equivalents) and dividing by the total amount of AI compute we estimate was delivered to the world’s major customers 3 months earlier, assuming a 3-month deployment lag.
Source-stated
Epoch AI says much of Microsoft’s and Amazon’s compute is not yet captured, and public-cloud deployments are difficult to attribute and quantify.
while much of Microsoft’s and Amazon’s compute is not yet captured. Public-cloud deployments are harder to attribute and quantify because the same infrastructure may serve several customers.
Source-stated
Epoch AI reports that its coverage of Chinese data centers is relatively low and says it aims to improve it.
Our coverage of Chinese data centers is relatively low, at 9-31%, and we aim to improve it in the coming months.
Source and contributor context
Epoch AI · Publisher
The publisher is identified here. No individual author has been inferred.
Source media
Media retrieval status: unchecked.
Terms and references
Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.
AI compute
The computing resources, meaning chips, data centers and electricity, used to train or run AI models.
Epoch AI
Epoch AI studies trends in AI capability, training resources and the infrastructure behind them.
Editorial provenance
Version 1. Publication ID: e49ea7a0-e7df-4c2c-71ac-4464ae41d806.
Prepared by openai/gpt-6-luna; grounding review by anthropic/claude-opus-5.5.
Source check: passed by anthropic/claude-opus-5.5 on 2026-10-29T16:57:32.802Z. This checks fidelity to the source, not whether its claims are independently true.
Editorial selection: accept; reasons: bridge_admitted_by_model.
Observation: a0cf76d0278f63ff7ccf067ced8ea65745681feb98afb51b60cdeb8ee07c8acf. Source text hash: f629a9e6df7b08ebd85e8f75c98401931a6bc965c2149ca097d23118ea952ca9.
Revision and source-check history
Version 1 · 2026-10-29T16:58:17.174Z
Epoch AI expands its AI data center research
Epoch AI’s post published September 16, 2026, describes an expanded AI Data Centers explorer and a redesign for viewing facility data. The organization says its research covers an estimated 44% of global AI compute across 86 data centers, and includes satellite imagery and details such as compute capacity, chip types, capital costs, and construction status. The explorer lets users compare sites over time and filter or recolor charts by owner, primary user, or country. Epoch AI says its coverage estimate compares compute capacity at operational data centers with estimated compute delivered to major customers, assuming a three-month deployment lag. It also says the data captures most of Meta’s AI compute, while much of Microsoft’s and Amazon’s compute is not yet included. Epoch AI identifies public-cloud deployments as difficult to attribute and quantify because infrastructure may serve several customers. It reports relatively low coverage of Chinese data centers and says it aims to improve that coverage. The announcement describes a growing dataset and new ways to inspect individual sites; it does not establish how complete the estimates are across companies or regions.
Original source
Epoch AI · Article
Source details
Original source date: Sep 16, 2026 (2026-09-16T00:00:00.000Z).
Date basis: Date supplied by the publisher’s feed.
Added to Escape Meridian: 2026-10-29T16:58:17.174Z. First observed: 2026-10-18T02:47:41.814Z. Retrieved: 2026-10-23T21:15:38.106Z.
Limits
Text as retrieved on October 23, 2026; originally published September 16, 2026. The coverage figures are estimates based on Epoch AI’s stated method; the post notes gaps in company and regional coverage and difficulty attributing public-cloud infrastructure. The post describes the explorer and its data coverage but does not independently establish how complete its estimates are.
Open question
How will Epoch AI’s coverage estimates change as it captures more public-cloud and Chinese data center capacity?
Source-stated
Epoch AI says the explorer covers an estimated 44% of global AI compute across 86 data centers.
Today, our research covers an estimated 44% of global AI compute, spread over 86 data centers.
Source-stated
Epoch AI says facility records include satellite imagery and data on compute capacity, chip types, capital costs, and construction status.
For each site, we provide satellite imagery and detailed facility data, including compute capacity in H100-equivalents, chip type(s), capital costs, and construction status.
Source-stated
Epoch AI estimates coverage by comparing operational data center compute capacity with estimated compute delivered to major customers, assuming a three-month deployment lag.
We estimate coverage by adding up all the compute capacity of operational data centers (in H100-equivalents) and dividing by the total amount of AI compute we estimate was delivered to the world’s major customers 3 months earlier, assuming a 3-month deployment lag.
Source-stated
Epoch AI says much of Microsoft’s and Amazon’s compute is not yet captured, and public-cloud deployments are difficult to attribute and quantify.
while much of Microsoft’s and Amazon’s compute is not yet captured. Public-cloud deployments are harder to attribute and quantify because the same infrastructure may serve several customers.
Source-stated
Epoch AI reports that its coverage of Chinese data centers is relatively low and says it aims to improve it.
Our coverage of Chinese data centers is relatively low, at 9-31%, and we aim to improve it in the coming months.
Source and contributor context
Epoch AI · Publisher
The publisher is identified here. No individual author has been inferred.
Source media
Media retrieval status: unchecked.
Terms and references
Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.
AI compute
The computing resources, meaning chips, data centers and electricity, used to train or run AI models.
Epoch AI
Epoch AI studies trends in AI capability, training resources and the infrastructure behind them.
Editorial provenance
Version 1. Publication ID: e49ea7a0-e7df-4c2c-71ac-4464ae41d806.
Prepared by openai/gpt-6-luna; grounding review by anthropic/claude-opus-5.5.
Source check: passed by anthropic/claude-opus-5.5 on 2026-10-29T16:57:32.802Z. This checks fidelity to the source, not whether its claims are independently true.
Editorial selection: accept; reasons: bridge_admitted_by_model.
Observation: a0cf76d0278f63ff7ccf067ced8ea65745681feb98afb51b60cdeb8ee07c8acf. Source text hash: f629a9e6df7b08ebd85e8f75c98401931a6bc965c2149ca097d23118ea952ca9.
passed · 2026-10-29T16:57:32.802Z · anthropic/claude-opus-5.5
Publication: e49ea7a0-e7df-4c2c-71ac-4464ae41d806
attribution — pass: The post is from Epoch AI's own site, and the reading consistently attributes the explorer's expansion, coverage estimates and methodology to Epoch AI.
We created the AI Data Centers explorer to make sense of this massive buildout
summary — pass: The title, summary and question accurately reflect the post: the expanded coverage, the redesign, the coverage method, company and regional gaps, and the stated aim to improve Chinese coverage. Terms such as 'AI compute' and 'public-cloud' are explained in context well enough for a general reader. The question follows from gaps the source names and is not leading.
Coverage varies by company: our data currently includes most of Meta’s AI compute
date — pass: The reading gives September 16, 2026, which matches the UTC calendar day of evidence.originalPublished. Nothing is presented as happening after that date, and in-text references such as 'by September 2026' are consistent with it. The evidence is a publisher page, so no archive disclosure is required.
we extrapolated that to 31.6 million by September 2026
support — pass: Each claim is backed by an exact source excerpt and attributed to Epoch AI. The limitations correctly note that the figures are Epoch AI's own estimates. The reading implies no unsupported benefits or costs.
Today, our research covers an estimated 44% of global AI compute, spread over 86 data centers.
media — pass: No media was selected and none was available among the verified candidates, so there is no identity or relevance issue.
Challenge or correct this reading
Send a correction or an opposing source with this reading’s URL. Disagreement and uncertainty can remain unresolved.