What are data centers?
A data center is a building — or a campus of buildings — filled with computers, often thousands of them, running nonstop, 24 hours a day.
It is the physical infrastructure behind the digital world. The emails you send, the videos you stream, the banking transactions you make, the weather forecasts you check, and the map apps you use all rely on servers housed on shelves inside real buildings.
In other words, the “cloud” is not abstract. It is physical, and it exists somewhere.
Most of those buildings have been invisible to most people for the past twenty years because they were small, scattered in cities, and didn’t make headlines. That has changed. The reason it has changed is artificial intelligence.
The AI difference
AI systems require enormous amounts of computation. Training a large AI model can consume as much electricity in a few months as several thousand homes use in a year. Running those models for everyone who wants to use them requires even more. To meet that demand, the companies building AI — Microsoft, Amazon, Google, Meta, Oracle, and many others — are building data centers far larger than anything that existed before.
There is also a physics reason for this scale. AI training is spread across thousands of chips that must constantly exchange data with each other, and even light itself takes measurable time to travel through fiber and switching equipment. At smaller scales, those delays are manageable. At frontier-model scale, they become a real performance penalty. That is why AI developers concentrate massive compute clusters inside hyperscale campuses with ultra-high-bandwidth, ultra-low-latency internal networks: the closer the machines are to one another, the faster they can train together.
The old data center was 20 to 50 megawatts of electricity demand — about the size of a small town. The new “hyperscale” AI data centers are 500 to 2,000 megawatts, with some proposed sites in Texas reaching 7,000 megawatts (7 gigawatts). For comparison, the largest single power plant in Texas before this wave was about 3,800 megawatts. Some of the new data center campuses, including the ones proposed in Hill County, build their own power plants on site to serve their own loads.
In one sentence — A modern hyperscale data center is the size of a town, runs continuously, and increasingly comes with its own dedicated power plant.
Why Texas
Texas has been a magnet for data center development for three reasons: land is available; the state grid (ERCOT) is independent and faster to interconnect to than most others; and Texas has been politically welcoming to large industrial customers. By early 2026, Texas had more than 400 data centers in operation or under construction — second nationally only to Virginia.
The current proposals in Hill County are part of that wave. Two are particularly large: a roughly 612-megawatt project near Hillsboro and a roughly 7.2-gigawatt project near Hubbard. Up to six other projects are reportedly in earlier stages of discussion.
What this brief does not address
This first brief explains what data centers are. The next briefs in the series cover how they affect water, power, noise, taxes, jobs, and rural character — and what the a Community Benefit Agreement could be structured to protect.
Next in this series
How do data centers affect the places where they operate? →
County Community Education Series · Prepared by Scope Technology and Manufacturing as advisor to Texas residents of unincorporated counties · May 2026