Sun Belt Scale: Phoenix and Dallas-Fort Worth

With the Ashburn frame in hand, walk the established roster. The point of this tour is not to memorize a ranking, which changes yearly, but to see that each primary market is a distinct answer to the power-on-a-date question, shaped by its utility, its regulator, and its geography. Start with the two that most resemble Ashburn in ambition: Phoenix and Dallas-Fort Worth.

Phoenix, meaning the suburban ring of Mesa, Chandler, Goodyear, and Buckeye more than the city core, became the runaway growth story of the American Southwest. The logic is almost embarrassingly clean. The land is flat, abundant, and cheap by coastal standards. The desert climate is hot but dry, and dry heat is friendlier to modern cooling than humidity, while the near-total absence of hurricanes, earthquakes, and floods makes underwriters comfortable signing long leases. Two utilities, Arizona Public Service and the Salt River Project, have been willing partners, and the state and localities offered tax treatment that welcomed the equipment. On top of that, Phoenix sits close enough to Southern California to serve as its lower-cost, lower-risk backyard, catching workloads that would rather not pay Los Angeles prices or sit on a fault line.

The obvious question in a desert is water, and it is a real one, but it is more nuanced than the headlines suggest. Newer facilities increasingly use air cooling and closed-loop systems that consume far less water than the evaporative designs of a decade ago, and operators have learned to make water commitments part of the entitlement conversation up front. Water is a gating item in Phoenix, not a disqualifier. The item that now gates hardest, as everywhere, is power: the same demand surge that strained Virginia is testing Arizona's utilities, and the interconnection timeline is the number a practitioner watches, not the thermometer.

Dallas-Fort Worth is the other high-growth giant, and it is best understood through one word: ERCOT. Most of Texas runs on its own electricity grid, largely walled off from the rest of the country, governed by the Electric Reliability Council of Texas. That isolation is a genuine double-edged sword. On the favorable edge, ERCOT is an energy-only market with a deregulated, competitive structure and a famously permissive posture toward building new generation and connecting new load; a developer can often move faster in Texas than almost anywhere else. On the unfavorable edge, that same grid delivered the catastrophic 2021 winter storm blackout, a reminder that speed and isolation carry reliability risk that customers price in. Around that grid, Dallas-Fort Worth offers cheap land, a large skilled labor pool, a central-time-zone location convenient to both coasts, and a business-friendly state government. Its natural hazard is severe weather, tornadoes and hail, which is a design cost rather than a deal-breaker.

It is worth dwelling on why the ERCOT structure produces such speed, because it explains a great deal about why so much new load, both crypto and AI, has landed in Texas. In most of the country a large customer must wait for a regulated, vertically integrated utility to plan, approve, and build generation and transmission through a multi-year regulatory process, and the customer joins a queue behind everyone else who had the same idea. ERCOT's competitive, energy-only design lets independent generators build where the money is and lets large loads contract directly in ways that can shave years off the timeline. The flip side is that no single entity is obligated to guarantee a reserve margin the way a traditional utility is, which is part of why the 2021 freeze cascaded the way it did. A practitioner reads that trade honestly: Texas offers speed and optionality in exchange for a reliability profile the customer must underwrite themselves, often by contracting for backup or by co-locating their own generation.

That distinction, between a market where you wait for the utility and a market where you can move on your own, matters more every year, because it previews the behind-the-meter tactics the emerging tier has taken furthest. Phoenix and Dallas both illustrate that even inside the primary roster, the same demand sorts itself by how each market answers the power question. A latency-sensitive cloud region that must sit near Southern California users leans toward Phoenix; a large training cluster that cares mostly about cost and speed to energization leans toward the Texas grid, where it can strike a faster deal. The workload chooses the market, and the deciding variable is almost always the shape and speed of the power arrangement, not the marketing brochure.

Set these two beside Ashburn and a pattern appears. Each is a large, liquid, by-now-proven market with a distinctive power story. Ashburn has the deepest liquidity and the tightest power. Phoenix trades desert-driven cooling efficiency and low disaster risk against a water conversation and a tightening grid. Dallas-Fort Worth trades regulatory speed and cheap inputs against the reliability question that ERCOT's isolation raises. A practitioner picking among them is not looking for the best market in the abstract. They are matching a specific workload's tolerance for latency, cost, and risk against the specific input profile of each place, and, first and last, checking which utility can actually deliver the megawatts on the date the business plan requires. The names on the roster are stable. The reason any given one wins a given deal is the timeline on its interconnection queue.