How Ashburn Became the Center of the Map

Start with a single county in Virginia, because everything about how this industry maps itself onto geography starts there. Loudoun County, and its neighbor Prince William, sit at the western edge of the Washington metro. The corridor is universally called Ashburn, or Northern Virginia, or by the people who pay the bills, simply the market. By any honest count it is the largest data center concentration on Earth, with more commissioned capacity than the next several North American markets combined and enough under construction to keep that lead for years. If you want to understand why one place becomes a primary market and another stays a footnote, you study Ashburn the way a geologist studies an outcrop: the whole history is exposed in the rock.

The accident that started it was fiber. In the 1990s one of the internet's principal exchange points, MAE-East, sat in nearby Tysons, and America Online was headquartered in Loudoun. Network builders laid conduit toward the traffic, and once fiber is in the ground it does not move. Every new route was cheaper to build alongside the existing ones, so the fiber thickened where it was already thick. That is the first law of market formation, and it is worth saying plainly: infrastructure compounds. A place that has a lot of something attracts more of that something, because the marginal cost of adding to a dense network is lower than the cost of starting a sparse one somewhere else.

Fiber brought the first data centers, the first data centers brought tenants, and tenants brought the thing that actually defines a primary market: liquidity. In a liquid market you can lease space on short notice, sell a building to a deep bench of buyers, hire a construction crew that has poured a hundred slabs exactly like yours, and call a utility account manager who has energized a dozen projects this quarter. None of that exists in a place with two data centers. Liquidity is why hyperscalers and enterprises pay a premium to be in Ashburn even when the raw inputs would be cheaper elsewhere. They are not buying cheap power. They are buying the certainty that they can execute, expand, and exit.

The counties, to their credit and eventual discomfort, made it easy. They zoned enormous tracts for data center use by right, meaning a developer could build without a discretionary rezoning fight, and they taxed the computing equipment inside those buildings at a rate that turned data centers into the fiscal engine of the county budget. By-right zoning plus a tax structure that rewards the use is a recipe for saturation, and saturation is exactly what arrived.

Which brings us to the constraint that now governs the whole story. For twenty years the scarce input in Ashburn was land, then it was skilled labor, and now it is unambiguously power. Dominion Energy, the utility serving the corridor, has publicly acknowledged that data center demand has outrun its ability to build transmission on the old timelines. Requests for large loads that once took a couple of years to serve began quoting into the late 2020s and beyond. The bottleneck is not generation in the abstract; it is the wires and substations that move electrons from where they are made to the specific parcel where a customer wants to plug in. Transmission takes many years to permit and build, and it cannot be conjured to match a construction schedule.

This is where the thesis of the whole course enters. In a mature primary market, the scarcest asset is not capital, which is abundant, nor land, which can usually be found. It is entitled, powered land: a parcel that is zoned for the use and, critically, has a firm, contracted path to enough electricity delivered on a known date. A raw field with a cheap price and a friendly county means nothing if the utility cannot energize it until the next decade. A parcel with a signed power commitment and a substation nearby is worth a fortune, because it collapses the single longest and least controllable line item on the project schedule. Everything else, a practitioner can buy, hire, or finance. Power on a timeline, they mostly cannot.

So the working definition of a primary market falls out of the Ashburn story. It is a place with dense fiber, deep operational liquidity, a proven regulatory and tax posture, and, above all, a track record of actually delivering power at scale, even if that delivery is now straining. Ashburn checks every box and is straining on the last one hardest, which is why the smart money spent the last few years asking a different question: if the original hub is running short of the one input that matters, where does the demand go next, and what will it take for the next place to earn the same trust? The rest of this course answers that question market by market, but hold onto the frame. Everything that follows is a variation on the search for entitled, powered land.