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 pricing, capacity, power, fabric, and latency across 240+ facilities and 40+ operators into a single queryable graph for AI agents, traders, and procurement teams.

[DC.COMPUTER] Layer 5 · Colocation

Switch

Every indexed Switch datacenter — facility-level capacity, GPU fleet, and country coverage.

Datacenter Portfolio

[AI Summary]

Switch runs 65 indexed datacenters across 2 countries, totaling 1,232 MW of IT capacity. The portfolio spans colocation tiers and deploys MI300X, H200, H100, A100, B200, L40S.

Facilities

1,232 MW

Total IT

Countries

GPU SKUs

[Facilities · 65]

Facility	Operator	Location	GPU	MW	PUE
Switch ASHBURN71	Switch	Ashburn, VA, USA	—	21	1.47
Switch ASHBURN72 South	Switch	Ashburn, VA, USA	MI300X	6	1.40
Switch ASHBURN76 Park	Switch	Ashburn, VA, USA	—	52	1.36
Switch ASHBURN79 East	Switch	Ashburn, VA, USA	—	54	1.35
Switch ASHBURN82 Park	Switch	Ashburn, VA, USA	—	9	1.40
Switch ASHBURN90 Central	Switch	Ashburn, VA, USA	—	16	1.43
Switch DALLAS23 Central	Switch	Dallas, TX, USA	—	38	1.31
Switch DALLAS103 East	Switch	Dallas, TX, USA	—	23	1.52
Switch CHICAGO34 North	Switch	Chicago, IL, USA	—	38	1.34
Switch CHICAGO37 Annex	Switch	Chicago, IL, USA	—	38	1.32
Switch CHICAGO41 Campus	Switch	Chicago, IL, USA	—	43	1.35
Switch CHICAGO44 Park	Switch	Chicago, IL, USA	H200	8	1.43
Switch CHICAGO48 Annex	Switch	Chicago, IL, USA	—	34	1.28
Switch LOS14 North	Switch	Los Angeles, CA, USA	—	24	1.23
Switch LOS21 West	Switch	Los Angeles, CA, USA	H100	17	1.43
Switch LOS23 East	Switch	Los Angeles, CA, USA	A100	38	1.35
Switch LOS25	Switch	Los Angeles, CA, USA	—	34	1.28
Switch LOS29	Switch	Los Angeles, CA, USA	—	32	1.24
Switch LOS30 Central	Switch	Los Angeles, CA, USA	—	12	1.47
Switch LOS34	Switch	Los Angeles, CA, USA	—	34	1.29
Switch LOS36 Central	Switch	Los Angeles, CA, USA	—	20	1.54
Switch LOS40 Campus	Switch	Los Angeles, CA, USA	—	12	1.48
Switch NEW54 East	Switch	New York, NY, USA	B200	8	1.41
Switch NEW58 East	Switch	New York, NY, USA	—	24	1.20
Switch NEW63 Annex	Switch	New York, NY, USA	MI300X	19	1.51
Switch NEW67 West	Switch	New York, NY, USA	—	37	1.34
Switch ATLANTA21 Annex	Switch	Atlanta, GA, USA	—	12	1.47
Switch ATLANTA25	Switch	Atlanta, GA, USA	B200	23	1.29
Switch PHOENIX22	Switch	Phoenix, AZ, USA	H200	16	1.49
Switch PHOENIX23 South	Switch	Phoenix, AZ, USA	L40S	3	1.40
Switch PHOENIX26	Switch	Phoenix, AZ, USA	H200	22	1.27
Switch DENVER1 Annex	Switch	Denver, CO, USA	—	18	1.23
Switch DENVER2 South	Switch	Denver, CO, USA	—	4	1.40
Switch DENVER3 Park	Switch	Denver, CO, USA	—	23	1.31
Switch DENVER5 West	Switch	Denver, CO, USA	—	19	1.21
Switch DENVER7 Campus	Switch	Denver, CO, USA	—	21	1.29
Switch DENVER9 West	Switch	Denver, CO, USA	H200	7	1.42
Switch DENVER10 Campus	Switch	Denver, CO, USA	—	26	1.35
Switch HOUSTON13 Central	Switch	Houston, TX, USA	—	14	1.52
Switch SAN2 Annex	Switch	San Antonio, TX, USA	—	20	1.30
Switch SAN9 Park	Switch	San Antonio, TX, USA	—	19	1.33
Switch BOSTON22 Annex	Switch	Boston, MA, USA	H100	2	1.39
Switch BOSTON23 Annex	Switch	Boston, MA, USA	—	7	1.46
Switch BOSTON30 Campus	Switch	Boston, MA, USA	—	14	1.22
Switch BOSTON31 Central	Switch	Boston, MA, USA	—	19	1.26
Switch COLUMBUS1 West	Switch	Columbus, OH, USA	—	3	1.51
Switch COLUMBUS4 Campus	Switch	Columbus, OH, USA	—	10	1.49
Switch COLUMBUS8 Annex	Switch	Columbus, OH, USA	—	21	1.31
Switch DETROIT1 East	Switch	Detroit, MI, USA	—	17	1.27
Switch KANSAS2 South	Switch	Kansas City, MO, USA	—	11	1.20
Switch KANSAS3	Switch	Kansas City, MO, USA	B200	17	1.31
Switch KANSAS7	Switch	Kansas City, MO, USA	—	3	1.42
Switch KANSAS10	Switch	Kansas City, MO, USA	MI300X	20	1.35
Switch KANSAS12 Campus	Switch	Kansas City, MO, USA	—	6	1.44
Switch SALT5 East	Switch	Salt Lake City, UT, USA	—	4	1.40
Switch SALT9 West	Switch	Salt Lake City, UT, USA	—	10	1.51
Switch JACKSONVILLE7 South	Switch	Jacksonville, FL, USA	—	7	1.47
Switch TORONTO10 West	Switch	Toronto, Canada	H200	15	1.21
Switch MONTREAL13 Campus	Switch	Montreal, Canada	B200	4	1.42
Switch MONTREAL20 Annex	Switch	Montreal, Canada	—	12	1.54
Switch MONTREAL21 Park	Switch	Montreal, Canada	—	27	1.47
Switch MONTREAL22 North	Switch	Montreal, Canada	—	17	1.23
Switch MONTREAL31 South	Switch	Montreal, Canada	—	17	1.23
Switch MONTREAL35 South	Switch	Montreal, Canada	MI300X	28	1.39
Switch MONTREAL39 South	Switch	Montreal, Canada	—	3	1.41

[FAQ]

How many datacenters does Switch operate?

65 indexed facilities totaling 1,232 MW across 2 countries.

Which GPUs does Switch deploy?

MI300X, H200, H100, A100, B200, L40S

In which countries does Switch operate?

Canada, USA

[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.