AI infrastructureUpdated 10 min read

The Neocloud Bible: How to Build an AI Data Center

AI data center build guide: 40-600 kW racks, $20M+/MW facility costs, liquid cooling, EU PUE 1.2 rules, and why modular builds win on speed.

Yuri MilyutinBy Yuri MilyutinManaging Partner, ModulEdge

Aisle of server cabinets inside a white modular data-center room, with the headline "How to build an AI data center."

An AI data center is a facility engineered around AI compute: its server and rack power, cooling interfaces and east-west network determine the design. Dense rack-scale training systems can require liquid cooling, while selected B200-class and RTX PRO Blackwell servers remain air-cooled. Power availability and delivery timing are major site constraints. Facility and IT costs must be compared on a stated scope; the cost table below separates those components.

This post covers what makes an AI data center different → the build sequence (power, cooling, network, structure) → cost and timeline benchmarks → EU compliance → why the build model itself is changing.

Why an AI data center is a different building

Most data centers were designed for a world that no longer exists. The Uptime Institute's 2025 global survey puts the mean rack density across the industry at 7.6 kW, with 5–9 kW the most common band. An NVIDIA DGX H100 node alone draws 10.2 kW. One server. More than the average rack.

And that was the gentle generation. A GB200 NVL72 rack runs at roughly 120–132 kW, and its GB300 successor runs 132–142 kW with ~155 kW peaks. The Rubin Ultra "Kyber" rack arriving in H2 2027 is specified around 600 kW. Per rack.

That's not an upgrade path. It's a different category of building — different power distribution, different cooling, different floor structure, different fire suppression assumptions. Operators who try to retrofit a 2015-era facility into hosting NVIDIA Blackwell-class hardware discover this in the engineering study, usually right before the project dies.

So the honest starting point for anyone asking "how do I build an AI data center" is this: stop thinking about it as a data center with more GPUs. Start thinking about it as an industrial power-and-cooling plant that happens to host compute.

Step 1: Secure power before anything else

Site selection used to start with fiber and land. Now it starts with megawatts.

In the US, projects that reached commercial operation in 2024 had spent an average of 55 months in the interconnection queue, according to Lawrence Berkeley National Laboratory's Queued Up report (2025). Four and a half years. In Europe it's worse: FLAP-D hub grid connections average 7–10 years (IEA, 2025), the Netherlands reports waits up to a decade, and Dublin has paused new data center connection agreements until 2028 entirely.

Which means your shortlist starts from substation capacity and queue position, not postcode. Increasingly it also includes generating your own power. Gas turbines, batteries, behind-the-meter renewables. The industry has shifted from "wait for the grid" to "bring your own electrons," a move we covered in detail in why data centers are done waiting for the grid.

One more AI-specific wrinkle most builders miss: step loads. Synchronized GPU training phases can swing facility draw 30–60% within milliseconds (McKinsey, 2025). Your UPS, harmonic filtering, and distribution have to be sized for the swing, not the average.

Step 2: Design the power architecture for the workload, not the habit

Traditional enterprise design defaults to 2N everywhere. AI changes the math.

GPU training clusters tolerate checkpoint-restart, so many operators now run N or N+1 on the compute blocks and reserve 2N for the network, storage, and control plane (Uptime Institute, 2025). Inference fleets serving production traffic push back toward full redundancy, because a dropped inference request is a dropped customer. The right answer depends on what the GPUs do, which is why redundancy topology should be a workload decision. We walk through the full N / N+1 / 2N logic in our modular data center power architecture guide.

For the gigawatt-scale Rubin era, NVIDIA and its OCP partners are pushing 800 VDC rack power distribution. If you're designing a facility today with a 10-year horizon, the DC-power question belongs in the base design review, not a future retrofit study.

Step 3: Cooling, where liquid is the baseline now

Air cooling has no single rack-kilowatt ceiling that applies to every server and cooling arrangement. Server fans, airflow, containment, inlet conditions and rear-door heat exchangers change the practical limit. The MDC500 air-cooled reference reaches up to 50 kW per rack, and NVIDIA offers air-cooled RTX PRO Blackwell servers. Rack-scale systems supplied for cold plates still need their supported liquid interface. Choose from the actual OEM requirements.

Direct-to-chip cold plates serve the liquid-cooled components in supported rack-scale systems such as GB200 NVL72. That interface brings requirements for CDUs or another supported hydraulic arrangement, facility loops, leak response and manifolds. It is not a universal requirement for every Blackwell system. Define the liquid and residual-air loads, OEM conditions and outdoor heat rejection together. Our liquid cooling guide for data centers covers the direct-to-chip vs immersion decision in depth.

There's a regulatory bonus here too. Liquid cooling delivers higher-grade, reusable heat, which matters enormously once German heat-reuse mandates apply (more on that below) and makes waste heat recovery a revenue conversation instead of a disposal one.

Step 4: Build the network for east-west traffic

AI traffic is GPU-to-GPU. Collective operations, all-reduce, parameter synchronization: the bandwidth that matters flows sideways between racks, not north-south to users.

The fabric market has flipped accordingly. When Dell'Oro Group began tracking AI back-end networks in late 2023, InfiniBand held over 80%. By Q3 2025, Ethernet took more than two-thirds of AI back-end switch sales, with 800 Gbps ports already the majority and 1,600 Gbps expected by 2027 (Dell'Oro, 2025). For the building, this translates into dense fiber pathways, tight rack adjacency to preserve copper NVLink domains, and physical room for rail-optimized topologies.

Step 5: Check the floor

A GB200 NVL72 rack weighs roughly 1.4–1.5 metric tons on a footprint of about 0.8 m². That's a point load near 1,875 kg/m², when typical raised floors are rated around 1,000 kg/m². The fix is reinforced slab-on-grade, which is exactly how purpose-built AI facilities and factory-built modules are constructed. A legacy raised-floor hall is, structurally, the wrong building.

Water deserves a line here too. A closed direct-liquid-cooling loop recirculates coolant, but the site can still consume water in cooling towers or adiabatic outdoor plant. Dry heat rejection can avoid routine process evaporation; the annual and peak water requirement depends on climate, temperatures and the complete heat path. A one-time loop fill is not an annual site-water budget.

What an AI data center costs

The numbers, with sources, because adjectives don't survive procurement:

ItemCostSource, year
Shell and core, global average$10.7M/MW (2025) → $11.3M/MW forecast 2026JLL, 2025–26
AI-optimized facility (liquid cooling, high-density electrical)$20M+/MWJLL, 2026
Tenant IT fit-out (GPUs, network, racks)up to ~$25M/MW additionalJLL, 2026
Chips vs facility, 1 GW AI campus~$20B chips vs ~$10B facility (~2:1)Epoch AI, 2025

Read that last row twice. The building is a minority of total capex, but it's the schedule gate for the majority. Every month of construction delay strands GPU capital worth roughly twice the building's cost. The cheapest thing you can buy in an AI data center project is time. Our modular data center cost breakdown shows where the facility money actually goes.

How long it takes, and why the answer is changing

A traditional hyperscale-class build runs 18–24 months of construction (IEEE Spectrum, 2025), but actual delivery now averages closer to four years once grid connection is included (IEA, 2025). Meanwhile the GPUs ship in weeks.

That mismatch is why the build model itself has flipped. McKinsey (2025) finds prefabricated and modular construction already makes up 40–60% of a typical data center project, and that standardized prefab builds cut delivery timelines by up to 50%. Factory-built modular data centers are a category-level speed advantage over traditional construction: the module is manufactured in parallel with site works and grid paperwork, then delivered with power, cooling, and fire suppression pre-integrated.

The hyperscale world has noticed. Microsoft's Fairwater facilities are standardized, replicated liquid-cooled designs stamped out as identical copies across sites. Crusoe, prime contractor on the 1.2 GW Stargate campus in Abilene, went further and opened a dedicated manufacturing plant in Colorado in March 2026 to mass-produce prefabricated modular AI factories. When the company building the world's largest AI campus decides the future is factory-built modules, the procurement signal is hard to miss.

When power is the bottleneck, the building must never be. That's the entire modular thesis, and the full modular data center guide covers how the delivery model works end to end.

The EU compliance layer

Building an AI data center in Europe means building into a reporting and efficiency regime that's already in force. Two instruments matter, and they are not the same thing.

The EU Energy Efficiency Directive (EED) is a transparency framework: under Delegated Regulation (EU) 2024/1364, any data center with ≥500 kW installed IT power must report energy use, PUE, WUE, heat-reuse factor, and renewables share annually to the European database. Reporting, not caps.

Germany's EnEfG is binding national law. Data centers commissioned on or after 1 July 2026 must achieve PUE ≤1.2 and reuse at least 10–15% of their waste heat, with the heat-reuse bar rising for later commissioning dates. Existing facilities face PUE ≤1.5 by July 2027 and ≤1.3 by July 2030.

A PUE target must be evaluated for the actual climate, load profile, redundancy state and measurement boundary. Neither air cooling nor modular construction establishes a universal annual PUE result, and liquid cooling does not guarantee one by design. The full regulatory picture is in our EU data center regulations guide.

Who's actually building all this: the neocloud context

The buyers driving AI data center demand increasingly aren't hyperscalers — they're neoclouds, GPU-first cloud providers built around AI workloads. CoreWeave guided to up to $35B of capex for 2026. Nebius is building greenfield capacity across Finland, the US, and Western Europe. Crusoe builds where the power is and manufactures its own modules.

The competitive logic across all of them, as one industry earnings roundup put it in 2026, has shifted "from GPU race to power wars." Advantage belongs to whoever has secured megawatts and can energize them fastest. How that business actually makes money (pricing, contracts, unit economics) is its own story, and we tell it in how the neocloud business works. And if you've already got the GPU allocation and need the path from loading dock to production, start with so you have GPUs, then what?

The build sequence, compressed

Power first: usable capacity, queue position or on-site generation, and the workload’s dynamic demand. Then cooling: the supported air and liquid interfaces, complete outdoor heat path and heat-reuse opportunity. Then network and structure: the cluster topology, fiber pathways, loaded rack mass and access. Then compliance: applicable reporting and national requirements. At every stage, identify what must be accepted before the GPUs can operate.

If the required operating date cannot be met, revise the delivery plan before committing the hardware. The GPUs depreciate either way.

An AI facility brief starts with the hardware the operator intends to run: compute racks, network and storage, liquid and air loads, and the required availability. ModulEdge selects the infrastructure around that schedule and the site, with upstream power and external cooling scope stated in the proposal.

FAQ

What is an AI data center?

An AI data center is a facility designed around AI servers or rack-scale systems, including their electrical load, cooling interfaces, network fabric and operating requirements. Some configurations use dense air-cooled servers; others use direct liquid cooling with a residual-air path. The exact hardware, rather than the AI label alone, determines the facility.

How much does it cost to build an AI data center?

AI-optimized facilities cost $20M+ per MW for the building, with GPU and IT fit-out adding up to ~$25M/MW on top (JLL, 2026). At a 1 GW campus scale, chips cost roughly twice the facility: about $20B versus $10B (Epoch AI, 2025). Shell-and-core averages $10.7M/MW globally as of 2025.

How long does it take to build an AI data center?

Traditional construction takes 18–24 months, but total delivery averages closer to four years once grid connection is included (IEA, 2025). US interconnection queues average 55 months (LBNL, 2025); EU hub markets run 7–10 years. Factory-built modular data centers compress the construction portion by 40–60%, with some AI modules delivered in 3–9 months.

What rack density does an AI data center need?

Rack density depends on the exact server count, power limits and rack-scale architecture. A dense air-cooled server rack and a GB200/GB300 NVL72 rack have different facility requirements. Obtain the OEM’s complete rack input and cooling schedule; treat future-platform figures as roadmap targets and verify them before committing a facility design.

Do AI data centers need liquid cooling?

Some do. GB200 and GB300 NVL72 use supported direct-liquid interfaces with separate residual-air loads. Other AI servers, including selected RTX PRO Blackwell systems, are air-cooled. There is no universal 40 kW rule: confirm the OEM configuration, airflow or liquid conditions, complete heat-rejection system and the site’s operating cases.

What EU regulations apply to AI data centers?

The EU Energy Efficiency Directive requires annual reporting of PUE, WUE, energy use, heat reuse, and renewables share for any facility with ≥500 kW installed IT power (Delegated Regulation 2024/1364). Germany's EnEfG goes further: binding PUE ≤1.2 and ≥10–15% waste-heat reuse for data centers commissioned from 1 July 2026.

What is a neocloud?

A neocloud is a GPU-first cloud provider offering AI compute (GPU-as-a-Service) rather than general-purpose IT: companies like CoreWeave, Nebius, Lambda, and Crusoe. Neoclouds are among the fastest-growing buyers of AI data center capacity, and their main competitive constraint is secured power and speed to energization.

‍

Bring the rack list. Or the site plan.

You do not need a finished specification. Send what is known, and we will identify the closest modular data center product and the decisions still standing between it and the site.

Contact form

Preview build: this form is not connected yet, so nothing was sent.