Founding Backer Campaign: lock in WP Pro 4 for $300/yr for life. Ends Dec 31.
Our data centers
AI data centers are going up all over the country, and many communities are pushing back over electricity prices, water use and noise. Those concerns are fair. But "data center" covers very different things, and a small web hosting company like ours is nothing like a facility built to train AI.
≈ 1 home's worth of electricity today
Closed-loop cooling
~70% zero-carbon power, solar planned
Electricity
The International Energy Agency estimates that a typical AI-focused data center uses as much electricity as 100,000 households, and the largest ones under construction will use 20 times as much.[1] Today, our servers use about as much as a single large home.
Average electricity draw, shown as the number of typical U.S. homes it could power. Linear scale.
Web7 Technologies today
Our servers in Albuquerque
≈ 1 home
Drawn wider than true scale so it's visible
A typical AI-focused data center
IEA estimate
≈ 100,000 homes
Largest AI data centers under construction
IEA estimate · 20× a typical one
≈ 2,000,000 homes
0
500K
1M
1.5M
2M
At true scale, our bar would be about one two-millionth the length of the longest bar. It's drawn at the smallest visible width so you can find it.
View the data as a table
Facility
Electricity use, in homes
Source
Web7 Technologies today
≈ 1 (a single large home)
Our own usage
Typical AI-focused data center
≈ 100,000
IEA [1]
Largest AI data centers under construction
≈ 2,000,000
IEA [1]
What's in a name
There's no single official definition, and news coverage often uses these terms interchangeably. Here's the rough landscape, and where we fit in it.
HYPERSCALE
Very large facilities built by the biggest cloud and tech companies, commonly drawing 100 megawatts or more.
AI
Facilities built around dense clusters of GPUs to train and run AI models, sometimes called "AI factories." A typical one uses as much electricity as 100,000 households.[1]
GIGAWATT
The newest and largest AI projects, planned at a gigawatt or more. The IEA says the largest under construction will use 20 times as much electricity as a typical AI data center.[1]
US
Servers that host websites, email and databases for businesses. Our entire footprint today uses about as much electricity as a single large home.
Side by side
AI and hyperscale data centers
Web7 Technologies
What it's for
Training and running AI models, and massive cloud platforms
✓ Hosting websites, email and databases for businesses
Electricity
A typical AI-focused facility uses as much as about 100,000 homes. The largest under construction, about 20 times that.[1]
✓ About as much as a single large home today
Server density
A single AI server rack can need up to 142 kW. The industry's most common rack draws about 11 kW.[3]
✓ Standard web hosting servers, not dense GPU clusters
Cooling and water
Large data centers can use up to 5 million gallons of water a day, as much as a town of 10,000 to 50,000 people.[2]
✓ Closed-loop cooling, which recirculates the same coolant instead of continuously evaporating water
Noise
Some large campuses have drawn complaints about a constant, low-frequency hum from cooling equipment.[5]
✓ A small footprint of servers, not a campus of industrial cooling equipment
The local grid
In the mid-Atlantic PJM grid, capacity prices jumped from $28.92 to $269.92 per megawatt-day in one auction, and analysts point to data center growth as one important contributor.[4]
✓ A draw comparable to one household, with a plan to stop using the local utility entirely
Our commitments
We're small today, and we intend to grow responsibly as we add capacity in Albuquerque and our planned Twin Cities location.
COOLING
Our data center recirculates the same coolant through a closed loop, rather than continuously evaporating water to stay cool.
POWER TODAY
Our utility, PNM, generates about 70% of its power from zero-carbon sources, so most of the electricity we use is already clean.
POWER TOMORROW
We plan to add our own solar to reach 100%+ clean energy, then scale it until we draw nothing from the local utility, so our growth never pushes up our neighbors' rates.
WHAT WE BUILD
We host websites, email and databases for businesses. We don't build GPU clusters for training AI models.
Isn't every data center basically the same?
No. "Data center" can mean anything from a single room of servers to a campus drawing a gigawatt of power. The resource concerns in the news mostly involve large hyperscale and AI facilities, which are built at a completely different scale from a hosting company like ours.
Does Web7 run AI data centers?
No. We host websites, email and databases for businesses on standard servers. We don't build or operate GPU clusters for training AI models.
What about your planned Minnesota location?
Our planned Twin Cities location, targeted for late 2027, is a second site for hosting our customers' websites. It adds capacity and regional redundancy, so each location can act as a failover for the other. It's not an AI campus.
What does "one home" mean on your chart?
We compare average electricity draw to typical U.S. households. For reference, the average U.S. residential customer used about 865 kWh of electricity per month in 2024.[6] Today our servers use about as much as a single large home.
Where does your power come from?
Our utility, PNM, generates about 70% of its power from zero-carbon sources. Our goal is to add our own solar to reach 100%+ clean energy, and eventually to generate all of our own power.
Sources
[1] IEA, Energy and AI, executive summary (April 2025)
[2] Environmental and Energy Study Institute, "Data Centers and Water Consumption" (June 25, 2025)
[5] Potomac Local, "Data center noise still shakes homes in Great Oak" (June 11, 2025)
[6] U.S. Energy Information Administration, Today in Energy (May 12, 2025)
Figures about AI and hyperscale facilities are industry-wide estimates from the sources above. Figures about Web7 describe our own current usage.
