AI, As I Go

Infrastructure

How AI data centers get built

What makes these facilities physically different, who actually owns them, and why they keep ending up in places you wouldn't expect.

How a site gets picked A hand-drawn decision path for picking an AI data center site, in five steps. 1 Power first, it matters most: Meta's Hyperion campus went to rural Richland Parish, Louisiana, where Entergy plans 10 gas plants to supply it (3 approved so far). 2 Land: a project in Cheyenne, Wyoming has grown to 2.7 gigawatts, nearly triple the state's entire current electricity use, picked for power and land rather than a city. 3 Latency: training can go anywhere power is cheap, since delays up to 100 milliseconds are fine; inference answers live prompts, so it moves to smaller sites near users. 4 Climate and water: the Nordics pair cold air with hydro power, cutting cooling cost and water draw. 5 Tax incentives: Clayton County, Georgia landed a roughly $1 billion data center with a tax abatement in 2025. Fiber and land round out the decision. how a site gets picked Read: Power comes first ↓1 power first matters most Meta Hyperion → ruralRichland Parish, LA: Entergy plans 10 gas plants Read: Power comes first ↓2 land Cheyenne, WY: 2.7 GW nearly 3× the state's wholecurrent electricity use picked for power + land, not a city Read: Distance works differently depending on the job ↓3 latency training: anywhere cheap ≤ 100 ms is fine inference: near users live prompts can't wait →smaller sites, closer in Read: Climate, land, and incentives ↓4 climate & water Nordics: cold air + hydro power= less cooling cost& water draw Read: Climate, land, and incentives ↓5 tax incentives Clayton County, GA: ~$1B data center landed with a taxabatement (2025) then fiber + landround it out How a site gets picked A hand-drawn decision path for picking an AI data center site, in five steps. 1 Power first, it matters most: Meta's Hyperion campus went to rural Richland Parish, Louisiana, where Entergy plans 10 gas plants to supply it (3 approved so far). 2 Land: a project in Cheyenne, Wyoming has grown to 2.7 gigawatts, nearly triple the state's entire current electricity use, picked for power and land rather than a city. 3 Latency: training can go anywhere power is cheap, since delays up to 100 milliseconds are fine; inference answers live prompts, so it moves to smaller sites near users. 4 Climate and water: the Nordics pair cold air with hydro power, cutting cooling cost and water draw. 5 Tax incentives: Clayton County, Georgia landed a roughly $1 billion data center with a tax abatement in 2025. Fiber and land round out the decision. how a site gets picked Read: Power comes first ↓1 power first matters most Meta Hyperion → ruralRichland Parish, LA: Entergy plans 10 gas plants Read: Power comes first ↓2 land Cheyenne, WY: 2.7 GW nearly 3× the state's wholecurrent electricity use picked for power + land, not a city Read: Distance works differently depending on the job ↓3 latency training: anywhere cheap ≤ 100 ms is fine inference: near users live prompts can't wait →smaller sites, closer in Read: Climate, land, and incentives ↓4 climate & water Nordics: cold air + hydro power= less cooling cost& water draw Read: Climate, land, and incentives ↓5 tax incentives Clayton County, GA: ~$1B data center landed with a taxabatement (2025) then fiber + land round it out
How an AI data center site gets picked, step by step, with the examples from this page. The sections below go into each one. Click or tap a title or picture in the sketch to jump to its section.

A traditional enterprise server rack draws somewhere around 5–10 kilowatts. A modern AI rack running Nvidia's GB200 NVL72 draws roughly 120 kilowatts — 15 to 20 times as much in the same footprint. That jump is why liquid cooling, once a niche choice, has become close to mandatory: once a rack crosses roughly 50 kW, air alone can no longer carry the heat away fast enough. The internal networking is different too — GPUs need to talk to each other constantly during training, not just to the outside internet, which calls for a denser, faster fabric than a traditional data center ever needed.

Three ownership models coexist today. Hyperscalers build and own their campuses outright — Meta's Hyperion site in rural Richland Parish, Louisiana is backed by Meta's own capital and its own dedicated gas power plants, built through Entergy specifically to supply it as the campus grew into a buildout of more than $50 billion. Ten plants are on order: three were approved in 2025, and a final Louisiana vote on the other seven is expected in late 2026, possibly as soon as November 18. Colocation providers take a landlord approach instead: they build and operate the facility, then lease out rack space, power, and network connections to multiple tenant companies. A newer “neocloud” model skips leasing space altogether and rents out GPU compute capacity directly instead.

Meta (META) — owner Equinix (EQIX) — landlord Digital Realty (DLR) — landlord CoreWeave (CRWV) — neocloud

For a modern AI campus, finding enough cheap, available power now matters more than almost anything else in picking a site. Meta's Hyperion campus went to a rural Louisiana parish specifically because of power availability, with Entergy planning ten new gas plants to supply it (three approved so far); a separate large AI data center project in Cheyenne, Wyoming, which Google acquired from developer Crusoe in November 2026, has grown to 2.7 gigawatts — nearly triple the state's entire current electricity use. Both were picked for power and land, not proximity to any major city.

Training and inference have different latency needs, and that shapes where each one gets built. Training workloads can tolerate delays of up to 100 milliseconds between regions since GPUs are mostly talking to each other, not to end users — which is why training campuses can sit almost anywhere power is cheap. Inference, which answers a live user prompt, is latency-sensitive by nature, so it increasingly gets pushed toward smaller facilities closer to where people actually are; with inference projected to overtake training as the majority of AI compute by 2030, that's starting to pull data center siting in two different directions at once.

Cooler climates cut cooling costs and water draw, which is part of why Nordic countries have attracted large data center builds for years, pairing cold air with abundant hydroelectric power. Tax incentives can also tip the decision: Clayton County, Georgia approved a tax abatement that helped land a roughly $1 billion data center in 2025, while Chesterfield County, Virginia, has taken a more mixed approach — approving incentive agreements for some data center projects in 2025, then upholding a denial of a separate proposal in 2026. Land, fiber access, and tax breaks round out the calculus once power and climate have narrowed the options. A single hyperscale campus can span hundreds of acres, and competition for buildable land near power and fiber has driven up real estate and grid-connection costs in several regions.