Matthews Logo

Navigation Menu

The AI Effect on Market Rents: Data Center, Edge Facilities, and Rent Wars
The AI Effect on Market Rents: Data Center, Edge Facilities, and Rent Wars featured image

The AI Effect on Market Rents: Data Center, Edge Facilities, and Rent Wars

The surge in the implementation of artificial intelligence (AI) is not just transforming technology, it’s rewriting the rules of industrial real estate. As hyperscale data centers and edge computing facilities multiply to meet AI’s insatiable appetite for computing power, they are driving up market rents in regions with robust energy and cooling infrastructure. The result is a new era of “rent wars,” where landlords are increasingly favoring server racks over forklifts and traditional tenants are being priced out.

 

Recent analysis from the National Association of Manufacturers (NAM) highlights how the cost to build and operate data centers has soared, with AI workloads requiring up to five times more power than traditional applications. This “AI effect” is most acute in power-rich regions, where access to reliable energy grids and advanced cooling systems is now the primary differentiator for landlords and tenants alike.

 

Power and Cooling: The New Currency of Industrial CRE

According to a recent Goldman Sachs report, the rapid rise in power density is disrupting the entire AI infrastructure ecosystem. Data centers that once required 10 to 20 megawatts are now demanding 100 megawatts or more, putting unprecedented pressure on local grids and utilities. In this environment, “it’s not just about square footage anymore,” as one Dallas property manager put it. “The first question we get from prospective tenants is, ‘How many megawatts can you deliver?’ If you can’t answer that, you’re out of the running.”

 

This shift is fueling a two-tiered market: properties with access to abundant power and cooling command premium rents, while those without are left behind. GlobeSt.com recently reported that some landlords are investing heavily in electrical upgrades and HVAC retrofits, hoping to attract AI and cloud tenants willing to pay double or triple the rates of traditional industrial users.

 

Mini-Case Studies: Phoenix, Northern Virginia, Quincy (WA)

Phoenix has emerged as a prime example of this trend. Once a logistics and light manufacturing hub, the city—especially the East Valley—now sees data center operators outbidding traditional tenants. Phoenix’s access to renewable energy and a favorable regulatory climate have made it a magnet for hyperscale facilities, with landlords increasingly opting for data center conversions to capitalize on higher rent premiums.

 

Northern Virginia is home to the world’s largest concentration of data centers, where competition for space and power is fierce. The region’s “Data Center Alley” is experiencing such intense demand that power constraints, not land, have become the primary bottleneck. As a result, landlords are prioritizing data center conversions over industrial tenants, fundamentally altering the region’s industrial mix.

 

Quincy, Washington leverages abundant hydroelectric power to attract cloud and AI infrastructure. The influx of data center development has driven industrial rents to record levels, forcing out agricultural and warehouse users who can’t compete with the lease rates offered by tech giants.

 

Traditional Industrial Users: Squeezed Out

The surge in AI-driven rents is squeezing out traditional industrial users—manufacturers, logistics firms, and small distributors—who simply can’t match the lease rates data center operators are willing to pay. As Jeremy Mercer, CEO of Matador Realty, explained on the Matthews Podcast, these tenants are often owner-users—local decision-makers running businesses like metal fabrication shops, screen printing operations, or small distribution companies. They typically need to buy or lease space for their day-to-day operations. “It was very, very relationship driven,” Mercer noted, emphasizing that for many of these business owners, “they make one deal in 10 to 20 years.”

But those relationships are now being tested as global tech companies outbid local businesses. As Mercer observed, “If you’re running a steel shop or a screen printing business, you can’t compete with a hyperscaler offering double or triple the rent.”

 

The New Reality: When Server Racks Outbid Forklifts

The phrase “when server racks are more profitable than forklifts” perfectly encapsulates this transformation. Hosting AI server racks, with their relentless demand for compute power, has fundamentally altered landlord incentives. Industrial real estate, once dominated by forklifts and pallets, is being reimagined as digital infrastructure. GlobeSt.com notes that this is leading to a bifurcation of the market: “The winners will be those with the foresight—and the infrastructure—to meet the demands of the digital age.”

 

Looking Ahead: The Bifurcation of Industrial CRE

As AI infrastructure continues to scale, expect further divergence in the industrial real estate market. Regions with access to cheap, reliable power and cooling will see continued rent escalation and intense competition, while traditional users will be forced to seek alternatives or relocate. The “rent wars” sparked by AI are only beginning, and the winners will be those who can deliver the power, cooling, and connectivity that tomorrow’s digital economy demands.

Similar Articles

The New Underwriting Playbook: What’s Driving Multifamily Decisions for 2026

Read More
EV Charging and the Second Life of Obsolete Gas Stations image

EV Charging and the Second Life of Obsolete Gas Stations

Read More
Midwest Self-Storage: Steady Hands Heading Into 2026 image

Midwest Self-Storage: Steady Hands Heading Into 2026

Read More
San Jose, CA Industrial Market Report Q1 2026 image

San Jose, CA Industrial Market Report Q1 2026

Read More