Why Data Centers Are the New Infrastructure Gold Rush

The last time American states competed this aggressively for a single category of infrastructure investment, it was automotive plants. Before that, steel mills. Now it's data centers — and the scale of capital chasing these projects makes the auto wars look quaint.

BlackRock and Nvidia recently led a $40 billion acquisition of a major data center firm. That single deal tells you everything you need to know about where institutional money thinks the next decade is headed.

The Surge Behind Data Center Development

Data centers aren't new. What's new is the demand curve — and it's nearly vertical.

Generative AI workloads require orders of magnitude more compute than traditional cloud applications. A single large language model training run can consume as much electricity as a small town uses in a month. That appetite for compute translates directly into demand for physical facilities: buildings full of servers, cooling systems, backup power, and fiber — lots of fiber.

The result is a construction boom that's less about square footage and more about power capacity, with modern hyperscale facilities routinely targeting 100MW, 200MW, even 500MW of connected load.

The key players shaping this market aren't just the obvious hyperscalers — Amazon Web Services, Microsoft Azure, Google Cloud. Specialized data center REITs like Equinix and Digital Realty have built global portfolios worth tens of billions. Private equity is funding merchant operators. Now sovereign wealth funds, pension funds, and infrastructure-focused asset managers like BlackRock are treating data centers as core infrastructure — the same category as toll roads and airports.

That institutional reclassification matters. When patient capital with 20-year time horizons starts underwriting an asset class, it signals something fundamental has shifted about how markets perceive long-term risk and return.

State Competition: Incentives, Location, and Power

Here's the dynamic that most coverage misses: states aren't just offering tax breaks. They're competing on fundamentals that are genuinely hard to replicate — available land, grid capacity, water access, permitting speed, and fiber connectivity.

Virginia's data center corridor in Northern Virginia — often called "Data Center Alley" — handles an estimated 70% of the world's internet traffic on any given day. That concentration exists because of compounding advantages: proximity to federal agencies, early infrastructure investment, and a permitting environment that (for years) moved fast. But that dominance is now being challenged precisely because Virginia's grid is straining and power costs are climbing.

States that can credibly offer 500MW+ of available grid capacity within 18 months of a commitment letter are winning deals that would have automatically gone to established markets five years ago.

Georgia, Texas, Ohio, Indiana, and the Carolinas are all aggressively courting data center development with packages that combine property tax abatements, accelerated depreciation on equipment, sales tax exemptions on servers and cooling hardware, and workforce development commitments. Some states are going further — pre-permitted industrial sites, dedicated utility interconnection queues, and even customized rate structures negotiated directly with rural electric cooperatives.

The incentive arms race is real. But sophisticated operators evaluate these packages carefully because a tax abatement that expires in year five doesn't help you amortize infrastructure with a 15-year useful life.

The Economic Case: Jobs, Tax Base, and Multiplier Effects

A single large data center — say, a 100MW campus — represents somewhere between $800 million and $1.5 billion in construction investment, depending on tier specification and redundancy design. That construction phase generates significant short-term employment: electricians, structural engineers, mechanical contractors, commissioning specialists.

The long-term picture is more nuanced. Data centers are not labor-intensive operations. A facility that cost a billion dollars to build might employ 50 to 150 people once operational. Critics of data center incentives routinely highlight this ratio, and they're not wrong to flag it.

What they often undercount is the tax base impact and the indirect economic activity — local suppliers, maintenance contractors, security firms, and the downstream effect of attracting technology-adjacent employers to a region.

Ohio's experience is instructive. The state's data center boom, anchored by major investments from Amazon and Google in the Columbus metro area, has catalyzed broader technology sector growth. Real estate values around these campuses have increased. Community colleges have launched technical training programs. The data center itself may employ 100 people, but the ecosystem it anchors can be substantially larger.

Infrastructure investment of this scale also upgrades the underlying grid and connectivity fabric of a region — improvements that benefit manufacturers, hospitals, universities, and every other power-intensive institution in the area.

The Real Constraints: Power, Water, and Regulatory Reality

The honest conversation about data center development has to include the constraints, because they're significant and getting more acute.

Power is the binding constraint for most markets right now. Utilities are reporting interconnection queues measured in years, not months. A developer who secures land today in a constrained market might not achieve commercial operation for four to six years — a timeline that makes financial modeling genuinely difficult. The solution increasingly involves developers funding their own transmission upgrades, negotiating directly with regional transmission operators, or co-locating with generation assets.

That last option — building a data center adjacent to or powered directly by a dedicated generation source — is driving serious interest in nuclear, particularly small modular reactors (SMRs). The appeal is obvious: carbon-free, baseload power with a footprint that can be sited near the load. The challenge is that SMRs at commercial scale remain years away from broad deployment, and the permitting pathway through the Nuclear Regulatory Commission is still being defined.

Water is the second major constraint. Evaporative cooling systems — still the dominant technology for large facilities — can consume millions of gallons per day. In water-stressed regions like the American Southwest, that's a genuine community conflict waiting to happen. Operators who deploy air-side economization, liquid cooling, or immersion cooling systems are gaining a real competitive advantage in water-constrained markets, not just a marketing talking point.

Regulatory hurdles vary dramatically by jurisdiction. Some states have streamlined permitting for infrastructure investment; others still require multi-year environmental review processes that make projects economically marginal before they break ground.

What Comes Next

The next decade in data center development won't look like the last one. A few trends are worth watching closely.

Energy efficiency is becoming a first-order competitive factor, not a sustainability checkbox. The metrics that matter — power usage effectiveness (PUE), water usage effectiveness (WUE), and increasingly carbon usage effectiveness (CUE) — are now being contractually specified by hyperscale tenants. Operators who can't demonstrate efficient design are getting cut from shortlists.

Geographic distribution is accelerating. The hyperscalers are deliberately spreading capacity across more regions to reduce concentration risk, satisfy data sovereignty requirements in international markets, and chase cheaper power. That means secondary and tertiary markets that can demonstrate grid reliability and permitting competence are genuinely in the game.

The AI infrastructure buildout is still in early innings. Training large models is compute-intensive, but inference — running those models at scale for billions of users — may ultimately require more total capacity than training does. That inference demand is sticky and grows with adoption, which means the capital commitment to data center infrastructure isn't a one-time surge. It's a structural, multi-decade investment cycle.

For states, developers, utilities, and landowners: the window to position for this cycle is open now, but it won't stay open indefinitely. Markets that solve the power problem first will capture a disproportionate share of investment. Those that don't will watch the deals flow to whoever got their grid right.

The infrastructure gold rush metaphor is apt — but unlike the original, the claims that matter here aren't staked with a pickaxe. They're staked with a signed utility interconnection agreement.

[INTERNAL LINK: data center investment trends]

[INTERNAL LINK: economic impact of data centers]

[INTERNAL LINK: state incentives for data centers]

Ready to explore the booming world of data centers? Discover more at InfraSale Marketplace.