Skip to main content
← Back to all posts

AI in Real Estate · Jul 7, 2026 · 9 min read

Edge Data Centers: The 40-Acre Footprint Quietly Reshaping Central Kentucky

The data-center news that makes headlines tends to come in gigawatts and billions. Hyperscale campuses spanning hundreds of acres. Hundred-megawatt first phases ramping to one gigawatt. Single deals reshaping a county’s tax base. That kind of project commands attention because the numbers are large and the announcements come with named tech companies attached.

There is a quieter version of the same story, and for central Kentucky, it may be the more important one. The category is called the edge data center, and the typical footprint is roughly one-tenth the size of a hyperscale build — 40 to 100 acres rather than 300 to 1,000, 80 to 100 megawatts rather than 500 to 1,000. Edge data centers do not generate the same headlines. They are also where most of central Kentucky’s actual data-center opportunity lives.

What an Edge Data Center Actually Is

The phrase “edge data center” has been used loosely enough to cover everything from a single rack in a cell-tower closet to a 36-megawatt regional campus. The useful definitions come from the industry analysts. The Uptime Institute describes edge as smaller facilities positioned closer to end users than centralized public cloud — designed to process data locally so latency drops from the 20-to-50-millisecond range of distant hyperscale into single-digit milliseconds. Gartner’s framing is similar: distributed computing and storage at the network edge, sized for real-time workloads that cannot tolerate round-trip latency to a hyperscale region.

The practical specs cluster in a recognizable range:

  • Power: 500 kilowatts to 2 megawatts for micro-edge facilities; 80 to 100 megawatts ramping to 300 or 400 for regional-edge campuses.
  • Land: 40 to 100 acres typical for regional edge, much smaller for micro.
  • Square footage: 10,000 to 50,000 square feet for regional edge.
  • Distance from MSA core: 20 to 50 miles, optimal range.
  • Latency target: single-digit milliseconds for inference and real-time workloads.
  • End-use mix: AI inference (regional model serving), 5G mobile-edge aggregation, content delivery network caching, industrial IoT and manufacturing automation, regional cloud nodes like AWS Wavelength and Microsoft Azure Stack Edge.

That sizing is not arbitrary. It is what the workload requires. AI inference at 10 to 15-plus kilowatts per rack does not fundamentally need a 500-megawatt campus — it needs nearby compute close to the population it serves. The same logic applies to 5G aggregation, CDN caching, and most non-training workloads.

Why Edge Is Growing Faster Than the Headlines Suggest

The global edge data-center market is forecast to grow from roughly $40 billion in 2026 to $105.8 billion by 2033, a 14.9 percent compound annual growth rate. Synergy Research projects hyperscale will still account for around 60 percent of global capacity by 2030, but edge is growing faster proportionally, and the underlying capex behind edge expansion runs into the hundreds of billions.

For tier-2 and tier-3 U.S. markets, the analyst consensus is that 300 to 400 megawatts of edge capacity will deploy by 2030 — distributed across Cleveland, Jacksonville, Minneapolis, Salt Lake City, St. Louis, Kansas City, Plano, and a long tail of similarly-sized metros. Central Kentucky belongs in that conversation. The reason is structural.

The tier-1 markets — Northern Virginia, Dallas-Fort Worth, central Ohio, Phoenix, Atlanta — are running into hard power constraints. Cushman & Wakefield reported in February 2026 that grid interconnection timelines in established hubs now routinely stretch to two and four years. Land costs and utility queue length push operators toward markets where the next megawatt is actually available, not promised. Tier-2 markets are typically 30 to 60 percent cheaper on capex per megawatt than mature markets — partly because land is cheaper, partly because utility coordination is less complex, and partly because permitting is faster.

Central Kentucky has the three things that combine into that lower-cost-per-megawatt math: abundant industrial-grade acreage, utility cooperatives that are increasingly responsive to industrial-scale inquiries, and a 30-to-50-mile ring around three meaningful MSAs (Louisville, Lexington, Bowling Green) plus a fourth at Owensboro.

Central Kentucky’s Edge Geography

If the optimal edge siting range is 20 to 50 miles from an MSA core, the map of central Kentucky becomes interesting. Around the Louisville MSA, the ring runs through Bullitt, Shelby, Spencer, and parts of Henry — Oldham, also in range, has moved in a restrictive zoning direction. Around the Lexington MSA, the ring captures Jessamine, Madison, Boyle, Mercer, and Scott. The Bowling Green MSA ring extends into Edmonson, with the Warren County core itself now restricted. Owensboro’s edge runs into Hancock and Ohio counties on one side and McLean on the other; Daviess proper has moved restrictive.

The Elizabethtown and Fort Knox area sits in its own category — Hardin, LaRue, and Meade are all in commercial range of Louisville’s southern edge while anchoring the Elizabethtown–Fort Knox metropolitan statistical area (MSA) in their own right. Hardin in particular benefits from the Glendale infrastructure spillover. For more on that corridor, see our deeper look at I-65 and the Glendale Megasite.

What that produces, in net, is a tractable list of central Kentucky counties with the geometric profile edge developers are searching for. Most of them have parcels in the 40-to-200-acre range. The key question is which of those parcels have the right power and zoning posture.

The Power Reality for Edge

Edge does not need hyperscale-class transmission, but it does need real power. Eighty to 100 megawatts available now — not promised in 2029 — is the threshold that turns a parcel into a candidate. That means substation proximity is the variable that matters most.

The central Kentucky utility mosaic is a mix. LG&E and KU cover most of Bullitt, Hardin, LaRue, Meade, Nelson, and points adjacent. Nolin RECC, a Touchstone Energy partner, fills in cooperative service across Hardin, LaRue, Meade, Breckinridge, Grayson, and Hart. Farmers RECC covers Hart and parts of Barren. Warren RECC plus TVA serves the Bowling Green area, where TVA brought the Transpark East 161 kV substation online in summer 2025. East Kentucky Power Cooperative supplies the eastern half of the state, including the Boyle, Mercer, Madison, and Jessamine area cooperatives.

Two practical implications follow. First, the substation map is the actual map. A parcel that is two miles from an existing 138 kV or 161 kV substation is a different asset than one that is ten miles away. Second, the cooperatives have begun engaging seriously with industrial-scale inquiries because cooperative service territories are increasingly the bottleneck for the state’s data-center pipeline. A relationship with the right key-accounts contact at a local cooperative can shorten the gap between “interesting parcel” and “qualified site” by months.

Fiber, the Edge DC’s Other Lifeline

Edge facilities cannot work without diverse fiber — minimum two carriers, preferably with route diversity in opposite directions. Central Kentucky’s fiber posture is better than its agricultural geography suggests. The I-65 corridor carries Lumen’s national backbone, plus Comcast Business, Windstream Kinetic, AT&T, and Cox. KentuckyWired’s state-built middle-mile backbone reaches all 120 counties across roughly 3,400 miles, which gives carriers the redundancy they need to commit to enterprise SLAs in places that would not otherwise have them.

Latency to the major data-center hubs falls in commercially viable ranges from Louisville and Lexington. Approximately 9 to 10 milliseconds to Ashburn, 7 to 8 to Chicago, around 8 to Atlanta, and 17 to 18 to Dallas. For most regional edge workloads — inference for southeastern population centers, CDN caching for the southeast and midwest, 5G mobile-edge for the I-65 freight corridor itself — central Kentucky’s latency posture works.

HB 775 and the Smaller-Tier County Story

The Kentucky tax incentive everyone talks about is House Bill 775, enacted March 2025. The headline figure is the $450 million capital-investment minimum, which qualifies a project for up to 50 years of sales and use tax exemption on data-center equipment. That headline is correct for the largest counties — Jefferson, Fayette, and a handful of others over 100,000 population.

The structurally significant detail, particularly for edge developers, is what HB 775 does in everywhere else. The Stites & Harbison legal brief describes the tiered structure plainly:

  • Counties of 100,000+ population: $450 million minimum, up to 50 years exemption.
  • Counties of 50,000 to 99,999: $100 million minimum, up to 25 years.
  • Counties under 50,000 population: $25 million minimum, up to 25 years.

A typical edge data-center campus in the 80 to 100 MW range carries a capital cost in the range of $50 to $150 million, depending on cooling architecture, building specifications, and contracted backup infrastructure. That budget hits the $25 million threshold easily and clears the $100 million threshold in most builds.

Translated to the central Kentucky map, that means LaRue, Hart, Nelson, Meade, Boyle, Mercer, Jessamine, and similar sub-50K counties are all positioned where a single edge campus would qualify for the tax exemption. The bill was, in effect, designed to push edge-scale projects into smaller counties — a fact that has been substantially under-reported in the broader coverage of HB 775.

Recent Edge Builds Worth Watching

The edge category is moving fast enough that the comparable transactions are mostly 2025 and 2026. EdgeConneX and Lambda launched a 23-megawatt AI factory in Chicago in 2025, with ready-for-service targeted in 2026. The same partnership has a 30-megawatt-plus campus operational in Atlanta. Edged Energy launched a 36-megawatt waterless-cooling AI facility in Mesa, Arizona, in April 2026. Duos Technologies deployed 15 modular edge units across Chicagoland by end of 2025, with a target of 50 by end of 2026.

In Kentucky, the comparable footprints are larger because the announcements happen to be from operators in the bitcoin-mining-to-AI transition. TeraWulf closed in May 2026 on the Muskie Data Campus in Eastern Kentucky — 1 GW of total planned capacity at EastPark Industrial Park, in two 500-megawatt phases through 2030, tied to a Kentucky Power 345 kV substation. TeraWulf’s earlier Justified Data Campus in Hancock County is sited for 480 megawatts. Both projects illustrate the state’s rising profile, even if their scale is closer to hyperscale than edge.

What is missing from the central Kentucky map, so far, is the 40-to-100-acre edge campus type that the population-tier structure of HB 775 most directly incentivizes. That gap is the opportunity.

What This Means for Landowners and Investors

The land conversation around edge data centers inverts the conventional industrial real-estate logic. For most heavy industry, bigger is better — more acreage, more flexibility, easier expansion path. For edge data centers, the right 50 acres beats the wrong 500 acres every time. Power posture, fiber proximity, MSA distance, zoning openness, and incentive-tier qualification matter individually and in combination.

A 50-to-100-acre parcel within two miles of an existing substation, on a fiber-served county road, in a sub-50K county that has not adopted DC-restrictive zoning, is now a categorically different asset than it was eighteen months ago. The market does not yet fully reflect that re-rating. Most owners of such parcels are not aware of what they hold. Most regional brokers are not tracking the variables that determine fit.

For developers, the implication is that the central Kentucky edge inventory is reachable through real conversations with landowners and cooperative-utility key-accounts contacts — not through searching active listings. Active listings will not include the parcels that fit best. The pre-positioning work has to happen now.

The Action Advisors team works central Kentucky’s I-65 and adjacent corridors every week. If you own, manage, or are evaluating a 40-to-200-acre parcel in the MSA-adjacent ring around Louisville, Lexington, Bowling Green, or Owensboro, the first conversation costs nothing and can usually tell you within an hour whether the variables align. Start with the I-65 corridor commercial inventory, or for Hardin County specifically, use this Hardin County search. For Elizabethtown-area commercial parcels, try this filter. For Shepherdsville and the northern end of the corridor, start here.


Grayson Bryan is a commercial and residential agent with Action Advisors and eXp Realty in Elizabethtown, Kentucky. The Action Advisors team works the I-65 corridor between Louisville and Bowling Green.

Related reading: Kentucky’s Data Center Boom: 30 Projects, $250M of Infrastructure, and a New State Law | Inside the I-65 Corridor: Glendale & Hardin County’s Data Center Story | Data Centers vs. Manufacturing: The Question Kentucky Communities Should Be Asking