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.

[D0] Discovery · Directory

Data centre directory · 833 sourced facilities

Every individually sourced facility record in the index, searchable and filterable. All filters and pages are plain links, so the full dataset is crawlable without JavaScript.

[AI Summary]

Datacenter.computer indexes 833 individually sourced data-centre facilities across 53 countries, 290 metros and 40 operators, with 4,536.77 MW of recorded capacity. Each record carries an operational status, capacity basis, verification status and confidence level; unknown fields are published as "Not publicly disclosed" rather than estimated. Modelled metro-level market signals are held in a separate research layer and never appear here as facilities.

660

Matching facilities

0 MW

Recorded capacity

With recorded MW

11 / 14

Page

Country

Metro

City

Operator

Facility type

Colocation 606 Hyperscaler 32 Sovereign 22

Operational statusreset

Operational 660

Capacity classificationreset

Not disclosed 660

Verification status

Operator Confirmed 435 Publicly Reported 225

[Results · 501–550 of 660]

Sourced data-centre facilities matching the current filters
Facility	Operator	Metro	Type	Status	Recorded MW	Confidence
EdgeConneX Boston	EdgeConneX	Boston, MA, USA	Colocation	Operational	Not publicly disclosed	High
EdgeConneX New York	EdgeConneX	New York, NY, USA	Colocation	Operational	Not publicly disclosed	High
EdgeConneX Richmond	EdgeConneX	Richmond, VA, USA	Colocation	Operational	Not publicly disclosed	High
EdgeConneX Norfolk	EdgeConneX	Norfolk, VA, USA	Colocation	Operational	Not publicly disclosed	High
EdgeConneX Cleveland	EdgeConneX	Cleveland, OH, USA	Colocation	Operational	Not publicly disclosed	High
EdgeConneX Columbus	EdgeConneX	Columbus, OH, USA	Colocation	Operational	Not publicly disclosed	High
EdgeConneX Indianapolis	EdgeConneX	Indianapolis, IN, USA	Colocation	Operational	Not publicly disclosed	High
EdgeConneX St. Louis	EdgeConneX	St. Louis, MO, USA	Colocation	Operational	Not publicly disclosed	High
EdgeConneX Toronto	EdgeConneX	Toronto, Canada	Colocation	Operational	Not publicly disclosed	High
EdgeConneX Amsterdam	EdgeConneX	Amsterdam, Netherlands	Colocation	Operational	Not publicly disclosed	High
EdgeConneX London	EdgeConneX	London, United Kingdom	Colocation	Operational	Not publicly disclosed	High
EdgeConneX Dublin	EdgeConneX	Dublin, Ireland	Colocation	Operational	Not publicly disclosed	High
EdgeConneX Munich	EdgeConneX	Munich, Germany	Colocation	Operational	Not publicly disclosed	High
EdgeConneX Madrid	EdgeConneX	Madrid, Spain	Colocation	Operational	Not publicly disclosed	High
EdgeConneX Lisbon	EdgeConneX	Lisbon, Portugal	Colocation	Operational	Not publicly disclosed	High
EdgeConneX Warsaw	EdgeConneX	Warsaw, Poland	Colocation	Operational	Not publicly disclosed	High
EdgeConneX Helsinki	EdgeConneX	Helsinki, Finland	Colocation	Operational	Not publicly disclosed	High
EdgeConneX Oslo	EdgeConneX	Oslo, Norway	Colocation	Operational	Not publicly disclosed	High
EdgeConneX Milan	EdgeConneX	Milan, Italy	Colocation	Operational	Not publicly disclosed	High
EdgeConneX Singapore	EdgeConneX	Singapore, Singapore	Colocation	Operational	Not publicly disclosed	High
EdgeConneX Johor	EdgeConneX	Johor Bahru, Malaysia	Colocation	Operational	Not publicly disclosed	High
EdgeConneX Mumbai	EdgeConneX	Mumbai, India	Colocation	Operational	Not publicly disclosed	High
EdgeConneX Chennai	EdgeConneX	Chennai, India	Colocation	Operational	Not publicly disclosed	High
EdgeConneX Tokyo	EdgeConneX	Tokyo, Japan	Colocation	Operational	Not publicly disclosed	High
EdgeConneX São Paulo	EdgeConneX	São Paulo, Brazil	Colocation	Operational	Not publicly disclosed	High
EdgeConneX Rio de Janeiro	EdgeConneX	Rio de Janeiro, Brazil	Colocation	Operational	Not publicly disclosed	High
EdgeConneX Santiago	EdgeConneX	Santiago, Chile	Colocation	Operational	Not publicly disclosed	High
EdgeConneX Buenos Aires	EdgeConneX	Buenos Aires, Argentina	Colocation	Operational	Not publicly disclosed	High
Ascenty São Paulo 1 (Cajamar)	Ascenty	Cajamar, Brazil	Colocation	Operational	Not publicly disclosed	High
Ascenty São Paulo 2 (Cajamar)	Ascenty	Cajamar, Brazil	Colocation	Operational	Not publicly disclosed	High
Ascenty São Paulo 3 (Cajamar)	Ascenty	Cajamar, Brazil	Colocation	Operational	Not publicly disclosed	High
Ascenty São Paulo 4 (Cajamar)	Ascenty	Cajamar, Brazil	Colocation	Operational	Not publicly disclosed	High
Ascenty São Paulo 5 (Cajamar)	Ascenty	Cajamar, Brazil	Colocation	Operational	Not publicly disclosed	High
Ascenty São Paulo 6 (Cajamar)	Ascenty	Cajamar, Brazil	Colocation	Operational	Not publicly disclosed	High
Ascenty Hortolândia 1	Ascenty	Hortolândia, Brazil	Colocation	Operational	Not publicly disclosed	High
Ascenty Hortolândia 2	Ascenty	Hortolândia, Brazil	Colocation	Operational	Not publicly disclosed	High
Ascenty Campinas 1	Ascenty	Campinas, Brazil	Colocation	Operational	Not publicly disclosed	High
Ascenty Campinas 2	Ascenty	Campinas, Brazil	Colocation	Operational	Not publicly disclosed	High
Ascenty Vinhedo 1	Ascenty	Vinhedo, Brazil	Colocation	Operational	Not publicly disclosed	High
Ascenty Vinhedo 2	Ascenty	Vinhedo, Brazil	Colocation	Operational	Not publicly disclosed	High
Ascenty Rio de Janeiro 1	Ascenty	Rio de Janeiro, Brazil	Colocation	Operational	Not publicly disclosed	High
Ascenty Rio de Janeiro 2	Ascenty	Rio de Janeiro, Brazil	Colocation	Operational	Not publicly disclosed	High
Ascenty Fortaleza 1	Ascenty	Fortaleza, Brazil	Colocation	Operational	Not publicly disclosed	High
Ascenty Fortaleza 2	Ascenty	Fortaleza, Brazil	Colocation	Operational	Not publicly disclosed	High
Ascenty Porto Alegre	Ascenty	Porto Alegre, Brazil	Colocation	Operational	Not publicly disclosed	High
Ascenty Jaguariúna	Ascenty	Jaguariúna, Brazil	Colocation	Operational	Not publicly disclosed	High
Ascenty Osasco	Ascenty	Osasco, Brazil	Colocation	Operational	Not publicly disclosed	High
Ascenty São Bernardo do Campo	Ascenty	São Bernardo do Campo, Brazil	Colocation	Operational	Not publicly disclosed	High
Ascenty Piracicaba	Ascenty	Piracicaba, Brazil	Colocation	Operational	Not publicly disclosed	High
Ascenty Sumaré	Ascenty	Sumaré, Brazil	Colocation	Operational	Not publicly disclosed	High

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.

Data centre directory · 833 sourced facilities

[Browse by country · 53]

Company → Datacenter → GPU → Customer → Industry