AI data center buildouts are reshaping capacity, energy, and infrastructure worldwide, with hyperscale projects increasingly tied to AI workloads. In the US, Northern Virginia leads hyperscale presence, while major cloud providers are scaling AI infrastructure through large capex programs. As electricity and water needs climb—from rising global power demand to cooling constraints—this page breaks down how capacity growth, GPU supply, and environmental limits shape where expansion lands.
Key Takeaways
- Global hyperscale data center capacity reached 44 GW in 2023, AI dominant
- US to add 5 GW data center capacity in 2024, 40% for AI
- Northern Virginia hosts 70% of US hyperscale capacity, AI hotspot
- Microsoft Azure capex $56B in FY2024, mostly AI data centers
- Amazon AWS invested $75B in capex 2024 for AI infrastructure
- Google Cloud capex $48B in 2024, driven by AI data centers
- Global NVIDIA GPU shipments for AI data centers reached 3.76 million units in 2023
- A typical AI training cluster uses 10,000+ NVIDIA H100 GPUs
- Microsoft Azure AI supercomputer features 10,000 GB200 GPUs
- Global data center electricity consumption reached 460 TWh in 2022, with AI workloads contributing significantly to growth
- By 2026, data center power demand could reach 1,000 TWh globally, driven by AI training and inference
- AI data centers in the US are projected to consume 35 GW of power by 2030, up from 3 GW in 2023
- AI data centers in Virginia consume 25% of state power, growing to 50% by 2030
- Microsoft data center in Iowa used 11.5 billion liters of water in 2022 for cooling AI servers
- Global data centers withdrew 1.13 trillion liters of water in 2021, with AI hyperscalers leading
AI-driven buildout is surging, pushing global data center power use toward 1,000 TWh by 2026.
Related reading
01 · Category
Construction And Locations23 stats
01
Global hyperscale data center capacity reached 44 GW in 2023, AI dominant
02
US to add 5 GW data center capacity in 2024, 40% for AI
03
Northern Virginia hosts 70% of US hyperscale capacity, AI hotspot
04
Microsoft plans 80+ new AI data centers by 2026
05
Amazon to build $100B in data centers over 15 years for AI
06
Google announces 25 new data center regions for AI expansion
07
Meta invests $10B in Louisiana AI data center campus
08
OpenAI partners for 5 GW new US data centers by 2028
09
Saudi Arabia launches 1.5 GW AI data center cluster
10
Europe data center pipeline hits 10 GW, AI driven
11
India plans 2 GW AI data centers by 2026
12
Oracle to deploy 100+ data centers globally for AI cloud
13
CoreWeave leases 1.3 GW across 28 sites for AI GPUs
14
Equinix xScale adds 1 GW for AI hyperscalers
15
Digital Realty pipeline 3 GW new capacity, 50% AI allocated
16
Switch builds 1 GW Citadel Campus in Nevada for AI
17
Vantage Data Centers raises $6.4B for 1 GW AI builds
18
NTT Global expands to 1.5 GW in US for AI colocation
19
China adds 500 MW AI data centers monthly in 2024
20
UAE launches 5 GW AI campus in Abu Dhabi
21
Flexential plans 500 MW AI-ready builds by 2026
22
Global data center construction spend hits $400B in 2024, AI fueled
23
AI data center construction costs rose 20% YoY to $12M/MW
Interpretation
Construction And Locations Interpretation
With hyperscale capacity hitting 44 GW globally in 2023 and the US adding 5 GW in 2024 with 40% earmarked for AI, construction and location decisions are clearly concentrating around AI hotspots like Northern Virginia, while major builders including Microsoft with 80+ new AI data centers by 2026 and Amazon’s planned $100B expansion over 15 years accelerate that trend.
02 · Category
Costs And Investments22 stats
01
Microsoft Azure capex $56B in FY2024, mostly AI data centers
02
Amazon AWS invested $75B in capex 2024 for AI infrastructure
03
Google Cloud capex $48B in 2024, driven by AI data centers
04
NVIDIA revenue from data center GPUs hit $26B in Q3 FY2025
05
Global AI data center market to $500B by 2030, CAGR 30%
06
Building 1 MW AI data center costs $10-15M
07
H100 GPU costs $30,000-40,000 each, key AI data center expense
08
Meta capex $37-40B in 2024 for AI compute and data centers
09
OpenAI valuation $157B, investing billions in custom AI data centers
10
CoreWeave raised $12B for AI data center expansion
11
Crusoe Energy secures $750M for AI cloud data centers
12
Lambda Labs funding $500M for GPU cloud data centers
13
Together AI raises $500M at $5.1B valuation for AI infra
14
xAI raises $6B for world's largest AI data center cluster
15
Oracle invests $10B in new AI data center campuses
16
Equinix $15B investment in xScale AI data centers
17
Digital Realty $7B JV for AI data center development
18
Vantage Data Centers $9.2B acquisition funding for AI builds
19
Global colocation market for AI $100B by 2028
20
Power costs for AI data centers average $0.07/kWh, 30% of opex
21
AI data center ROI projected 20-30% for hyperscalers by 2027
22
TSMC capex $30B/year building AI chip fabs for data centers
Interpretation
Costs And Investments Interpretation
Under the Costs And Investments category, the data shows a clear surge in funding for AI buildouts, with major cloud players alone projecting tens of billions in 2024 capex across Microsoft Azure at $56B, Amazon AWS at $75B, and Google Cloud at $48B, alongside a 30% CAGR to a $500B global AI data center market by 2030 and costs of roughly $10M to $15M per 1 MW, underscoring that AI infrastructure is becoming a top-level capital priority.
03 · Category
Hardware And Capacity24 stats
01
Global NVIDIA GPU shipments for AI data centers reached 3.76 million units in 2023
02
A typical AI training cluster uses 10,000+ NVIDIA H100 GPUs
03
Microsoft Azure AI supercomputer features 10,000 GB200 GPUs
04
xAI's Colossus cluster launched with 100,000 NVIDIA H100s, world's largest
05
Meta plans 350,000 NVIDIA H100 equivalents by end-2024 for AI training
06
Google DeepMind's TPU v5p pods scale to 8,960 chips for AI workloads
07
Amazon Trainium2 chips deliver 4x performance per watt for AI inference
08
Cerebras Wafer-Scale Engine WSE-3 has 4 trillion transistors for massive AI models
09
Grok-1 trained on 314B parameter model using custom GPU clusters
10
AMD Instinct MI300X offers 192 GB HBM3 for AI data centers
11
Intel Gaudi3 AI accelerator competes with 50% more throughput than H100
12
Tesla Dojo D1 chip has 50 petaflops FP16 for AI video training
13
Oracle OCI Supercluster supports 131,072 NVIDIA GPUs in one RDMA fabric
14
SambaNova SN40L chip enables 1.7 exaflops AI compute in rack
15
Graphcore IPU Colossus MK2 GC200 scales to 72 chips per card for AI
16
Huawei Ascend 910B delivers 60% HBM capacity of H100 for China AI centers
17
Tenstorrent Wormhole n300 has 128 cores for efficient AI inference
18
D-Matrix Corsair chip offers 10x better inference perf/Watt
19
Global AI accelerator market to ship 11M units by 2027
20
hyperscalers deployed 4M GPUs in 2024 for AI capacity
21
CoreWeave operates 250,000 NVIDIA GPUs across 32 data centers
22
Lambda Labs AI cloud has 20,000 H100s online
23
Together AI cluster with 20,000 H100s for open models
24
Crusoe Energy AI platform deploys 10,000 H100s on flared gas
Interpretation
Hardware And Capacity Interpretation
Hardware and capacity for AI data centers are scaling rapidly, with NVIDIA H100 demand growing from typical 10,000 GPU training clusters to mega deployments like xAI’s 100,000 H100 Colossus launch and Meta’s planned 350,000 H100 equivalents by end 2024.
More related reading
04 · Category
Power And Energy24 stats
01
Global data center electricity consumption reached 460 TWh in 2022, with AI workloads contributing significantly to growth
02
By 2026, data center power demand could reach 1,000 TWh globally, driven by AI training and inference
03
AI data centers in the US are projected to consume 35 GW of power by 2030, up from 3 GW in 2023
04
A single ChatGPT query requires 2.9 Wh of electricity, 10x more than a Google search
05
NVIDIA H100 GPUs in AI clusters consume up to 700W per chip, enabling massive power draw in hyperscale setups
06
Data centers accounted for 2% of global electricity in 2022, expected to rise to 3-4% by 2030 due to AI
07
Microsoft's AI data centers power usage grew 34% year-over-year in 2023
08
Google's data centers used 18.3 TWh in 2022, with AI optimization reducing PUE to 1.10
09
AI training for GPT-4 consumed energy equivalent to 1,000 US households for a year
10
By 2030, AI could drive US data center power demand to 8% of national total
11
Hyperscale AI data centers require 100-500 MW per facility
12
Electricity demand from AI data centers could increase 160% by 2030 per Goldman Sachs
13
Amazon AWS AI workloads increased power consumption by 20% in 2023
14
A 1 GW AI data center can power 750,000 homes
15
Meta's AI data centers target PUE under 1.10 with liquid cooling, consuming 500 MW+ per site
16
Global AI compute power demand to hit 85 GW by 2027
17
Training one large AI model like BLOOM uses 433 MWh
18
US data centers to consume 35 GW for AI by 2030, equivalent to 10 new nuclear plants
19
China's AI data centers power usage to triple to 200 TWh by 2027
20
Oracle's AI data centers plan 100+ facilities with 2 GW total power by 2028
21
AI inference power per query rising 50% annually
22
Europe's AI data centers to require 35 GW by 2030
23
Tesla's Dojo AI supercomputer cluster draws 100 MW
24
Worldwide data center power to reach 8% of global electricity by 2030 due to AI
Interpretation
Power And Energy Interpretation
Power and energy demand is set to escalate fast as global data center electricity use climbs from 460 TWh in 2022 to a potential 1,000 TWh by 2026, with AI workloads pushing data center electricity’s share of the world’s total from 2% to about 3 to 4% by 2030.
05 · Category
Water And Cooling25 stats
01
AI data centers in Virginia consume 25% of state power, growing to 50% by 2030
02
Microsoft data center in Iowa used 11.5 billion liters of water in 2022 for cooling AI servers
03
Global data centers withdrew 1.13 trillion liters of water in 2021, with AI hyperscalers leading
04
Google's data centers used 5 billion gallons of water in 2022, up 20% due to AI
05
A single AI data center can evaporate 1 million gallons of water per day for cooling
06
Meta's AI data centers in Arizona consumed 170 million gallons monthly in 2023
07
Cooling accounts for 40% of data center energy, critical for AI GPU clusters
08
Shift to liquid cooling in AI data centers reduces water use by 30% vs air cooling
09
OpenAI's US data centers projected to use 1 trillion gallons water over 5 years
10
Amazon's Virginia data centers used 6.75 billion liters water in 2022 for AI cooling
11
AI training clusters require advanced cooling, consuming 20-30% more water per MW
12
Ireland's data centers, many AI-focused, used 25% of national water despite 2% population
13
Direct-to-chip liquid cooling in NVIDIA DGX systems saves 50% water vs traditional
14
Global AI data center water demand to rise 50% by 2027
15
Microsoft's Santarém facility recycles 95% cooling water for AI ops
16
Hyperscale AI centers in arid regions face 20% higher water stress
17
Two-phase immersion cooling cuts AI data center water use by 90%
18
Chile's AI data centers strained Santiago's water supply by 20% in 2023
19
Rear-door heat exchangers reduce cooling water by 40% in AI racks
20
Global data center cooling water to hit 2.5 trillion liters by 2025, AI driven
21
Equinix AI facilities target zero-water cooling with dry coolers
22
AI GPU density requires 50 kW/rack cooling, upping water needs 4x
23
NVIDIA's GB200 NVL72 needs 120 kW/rack, demanding advanced water-efficient cooling
24
World's largest AI data center in Saudi Arabia uses seawater cooling to minimize fresh water
25
NVIDIA DGX H100 systems deploy with closed-loop liquid cooling to cut water 70%
Interpretation
Water And Cooling Interpretation
AI data centers are rapidly driving water demand for cooling, from Virginia’s power load rising from 25% now to a projected 50% by 2030 and Google’s water use jumping 20% in 2022 to 5 billion gallons, while single facilities can evaporate about 1 million gallons per day and global withdrawals reached 1.13 trillion liters in 2021.
Reference
Cite This Report
This report is designed to be cited. We maintain stable URLs and versioned verification dates. Copy the format appropriate for your publication below.
APA
Nathan Caldwell. (2026, February 24). AI Data Centers Statistics. Gitnux. https://gitnux.org/ai-data-centers-statistics
MLA
Nathan Caldwell. "AI Data Centers Statistics." Gitnux, 24 Feb 2026, https://gitnux.org/ai-data-centers-statistics.
Chicago
Nathan Caldwell. 2026. "AI Data Centers Statistics." Gitnux. https://gitnux.org/ai-data-centers-statistics.
Sources & references
73 datasets cited across this report · attribution is report-level
sustainability.aboutamazon.com
aws.amazon.comir.aboutamazon.com