What is an AI datacenter?

An AI datacenter is a facility purpose-built or retrofitted for GPU and accelerator workloads — typically at rack densities above 30 kW, with liquid-cooling readiness and high-bandwidth fabric (InfiniBand NDR, NVLink, or Spectrum-X). At least 40% of IT load runs AI training or inference on GPUs such as NVIDIA H100, H200, B200, or GB200 NVL72.

Which datacenter has the most GPUs?

As of 2026, xAI Colossus in Memphis, Tennessee holds the title of the world's largest single-site GPU deployment, with more than 200,000 NVIDIA H100 and H200 GPUs in one InfiniBand domain. Meta Hyperion (Louisiana) and OpenAI Stargate Abilene (Texas) are scaling toward gigawatt-class deployments that will overtake it within the next 18 months.

What is the largest AI cluster?

xAI Colossus is the largest contiguous AI training cluster in 2026, exceeding 200,000 NVIDIA H100/H200 GPUs on a single InfiniBand NDR 400G fabric. Meta's Hyperion build and OpenAI's Stargate program are constructing multi-site fabrics that will exceed one million accelerators across coupled campuses by 2027.

Which GPU cloud is cheapest for H100?

On-demand: Together AI ($1.75/hr) and Voltage Park ($1.89/hr) lead. Reserved 1-year: Voltage Park ($1.20/hr) and Crusoe ($1.39/hr) are typically lowest. AWS p5 sits at the top of the price stack at ~$4.10/hr on-demand.

How is Datacenter.computer different from a datacenter directory?

Datacenter.computer is an agent-readable intelligence layer, not a directory. It normalizes capacity, power, fabric and location across 833 sourced facilities and 40 operators into a single queryable graph for AI agents, traders and procurement teams.

[MAP] Discovery · Map

Global data centre map · 833 sourced facilities

A geographic view of the sourced index. The map is a visual aid only — the authoritative, crawlable dataset is the HTML table below it.

[AI Summary]

The map plots 833 individually sourced data-centre facilities across 53 countries and 290 metros. Coordinates are metro or campus-level approximations, never surveyed positions, and no marker is generated from a modelled market signal. The same records are listed as standard HTML links below so crawlers and agents can enumerate every facility without JavaScript.

Matching facilities

0 MW

Recorded capacity

With recorded MW

1 / 1

Page

Countryreset

Malaysia 7

Cityreset

Johor Bahru 7

Operator

AirTrunk 2 Chindata Group 1 EdgeConneX 1 Keppel Data Centres 1 Princeton Digital Group 1 STT GDC 1

Facility type

Colocation 4 Hyperscaler 3

Operational statusreset

Operational 7

Verification status

Operator Confirmed 5 Publicly Reported 2

[Results · 1–7 of 7]

Sourced data-centre facilities matching the current filters
Facility	Operator	Metro	Type	Status	Recorded MW	Confidence
Chindata Johor	Chindata Group	Johor Bahru, Malaysia	Hyperscaler	Operational	Not publicly disclosed	High
STT GDC Johor	STT GDC	Johor Bahru, Malaysia	Colocation	Operational	Not publicly disclosed	High
AirTrunk JHB1 Johor Bahru	AirTrunk	Johor Bahru, Malaysia	Hyperscaler	Operational	Not publicly disclosed	High
AirTrunk JHB2 Johor Bahru	AirTrunk	Johor Bahru, Malaysia	Hyperscaler	Operational	Not publicly disclosed	High
EdgeConneX Johor	EdgeConneX	Johor Bahru, Malaysia	Colocation	Operational	Not publicly disclosed	High
Keppel DC Johor 1	Keppel Data Centres	Johor Bahru, Malaysia	Colocation	Operational	Not publicly disclosed	High
PDG JH1 Johor	Princeton Digital Group	Johor Bahru, Malaysia	Colocation	Operational	Not publicly disclosed	High

[Map view]

[MAP] Global Datacenter Index

833 facilities · 40 operators · 53 countries · 4,536.77 MW

Showing top 250 of 833 facilities in this view.

Positions are approximate. Verify any siting decision against the operator source linked on each facility page.

Counts are computed live from the sourced facility index. Fields without a public basis show “Not publicly disclosed”. See the methodology, the data dictionary and the public API.

[Knowledge Graph]

Company → Datacenter → GPU → Customer → Industry

Every entity on this site is cross-linked. Follow the graph from operators down to specific facilities, GPU clusters, customers, and sectoral context.