The Ultimate Guide for Data Centers

Updated Leasing Volume

The U.S. data center leasing market set records across virtually every metric in 2025. Across the eight primary North American markets, users absorbed 2,497.6 megawatts of capacity, a 38% increase over 2024 and more than five times the pace of just three years earlier. Northern Virginia reclaimed the top position nationally with 1,102 MW of net absorption, more than doubling its 2024 total. Dallas-Fort Worth surged to 470.8 MW, up from just 46.8 MW the prior year. [10]

Hyperscale lease commitments moved at an even more dramatic pace. In Q3 2025 alone, 7.4 gigawatts of new hyperscale capacity was committed, more than the 7 GW leased in all of 2024. The momentum continued into Q4, with new leasing across top U.S. markets exceeding 1,600 MW in the quarter, led by Chicago, which saw an all-time high of more than 300 MW in new deals. Next-tier markets also posted impressive absorption, with 1.4 GW in combined new deals largely driven by neocloud providers. [11]

Market Rankings and Expansion

Northern Virginia remains the largest U.S. data center market by a commanding margin, expanding to 4,039.6 MW of total colocation inventory in 2025, a 37% increase year-over-year, as the region delivered more than one gigawatt of new capacity in a single year. The region has nearly three and a half times more data center capacity than all secondary U.S. markets combined. Colocation vacancy in Northern Virginia fell to 0.5%, with the majority of 2026 capacity already committed and preleasing extending into 2027 and beyond. [12]

Dallas-Fort Worth has become the third North American market to surpass 1 GW of total supply, joining Northern Virginia and Atlanta. Atlanta is one of the fastest-growing regions, with more than 2 GW currently under construction. The combination of AI training demand and faster long-distance fiber networks has opened the door for markets previously considered too remote for large-scale deployment. Nevada, Pennsylvania, and Michigan are all gaining traction due to available land, flexible permitting, and better access to power. [13]

Vacancy and Pricing

Despite a 36% year-over-year increase in total primary market supply to 9,432 MW, the national vacancy rate fell to a record low of 1.4% at year-end 2025. Average monthly asking rates for 250 to 500 kW requirements at N+1/Tier III rose 6.5% year-over-year to approximately $196 per kilowatt per month, the fourth consecutive annual increase. Rental rates are projected to continue rising at a mid-single-digit pace per annum through 2030, with hyperscale rent growth outpacing colocation. Pricing pressure has been most acute for large-scale deployments: rates for requirements of 10 MW or more jumped 12.5% year-over-year as competition for contiguous, power-ready capacity intensified. [14]

On the supply side, approximately 12 GW of new third-party landlord capacity is expected to be added across top U.S. markets through 2030, representing a compound annual growth rate of approximately 9-10%. Dallas, Columbus, and Atlanta are projected to see the highest supply growth rates. These estimates cover only established markets; actual industrywide growth will be significantly larger as next-tier locations absorb AI training deployments requiring immediate access to hundreds of megawatts of power. [15]

Power as the Primary Constraint

Power delivery timelines continue to shape every aspect of the market. New capacity under construction in primary markets declined for the first time since 2020 in the second half of 2025, dropping to 5,994 MW from 6,350 MW at year-end 2024. The decline is not a signal of weakening demand; it reflects the practical reality that many planned projects cannot advance until power infrastructure is upgraded or secured. For developers and investors able to navigate power procurement and secure grid interconnection, this constraint creates significant first-mover advantage.[16]

Updated Hyperscale Examples

The scale of capital investment and leasing commitments continues to accelerate. In 2025, the sector recorded its first single lease exceeding one gigawatt (1.4 GW in West Texas), and a $30 billion joint venture was formed to develop a 2 GW campus in Louisiana that is already fully leased. Leading hyperscalers are collectively projected to invest more than $700 billion in capital expenditure in 2026, representing roughly 60% year-over-year growth. [17]

A new user category is also reshaping demand. Neocloud providers, companies like CoreWeave, Crusoe, Lambda, and Nebius that offer GPU-as-a-Service to AI companies, have emerged as significant lessees, inking several gigawatts of new deals in 2025 and driving much of the absorption in next-tier markets. Their capacity requirements rival those of traditional hyperscalers and are adding a new dimension to the demand picture. [18]

Power and Cooling

A single AI query consumes approximately 2.9 watt-hours of electricity, compared with 0.3 watt-hours for a traditional web search, roughly ten times the energy intensity. When scaled across billions of daily interactions, this disparity is a key driver of the infrastructure expansion reshaping the data center sector. [19]

The rise of AI is forcing a fundamental shift in data center design. As power density at the rack level increases to support GPU-intensive training and inference workloads, traditional air-cooling systems are reaching their physical limits. The transition to liquid cooling is identified as one of four critical challenges facing data center production at scale, alongside power procurement, workforce capacity, and construction efficiency. [20]

Power Consumption Context

U.S. data centers consumed more than 4% of the nation’s total electricity in 2023, and a single large facility can consume as much electricity as 50,000 homes. Global data center electricity consumption reached approximately 415 terawatt-hours in 2024 and is expected to more than double to 945 TWh by 2030, slightly more than Japan’s total annual electricity consumption today. [21]

Water and Cooling Efficiency

As the industry scales, cooling technology is evolving rapidly. The sector is moving toward less water-intensive solutions, including air-cooled and closed-loop systems, that reduce environmental impact while supporting higher power densities. Water-efficient design and responsible sourcing are becoming increasingly important to both operational performance and community acceptance, particularly as local permitting processes incorporate water usage assessments. [22]

Sustainability

Green certifications and environmental performance are increasingly influencing where and how data centers can be built. Several states have introduced legislation in 2025-2026 conditioning new data center approvals on environmental impact assessments, grid readiness, or water usage studies. For operators and investors, this regulatory evolution reinforces the value of sustainable design, renewable energy procurement, and proactive community engagement, not just as reputational benefits, but as operational prerequisites. [23]

Why Institutional Capital is Flowing In

Record leasing volume, record-low vacancy, long-term lease structures backed by investment-grade credit, and critical infrastructure status have made data centers one of the most fundamentally sound real estate asset classes available. The sector’s investment profile has matured rapidly: transaction volume reached historic levels in 2025, complex capital structures including joint ventures, forward sales, and preferred equity are now standard, and record data center ABS issuance exceeded $17 billion. Development economics remain highly attractive, with profit margins of 50% or more in established markets. [24]

$1.5 billion+ in Lincoln’s investment portfolio across 21 data center investments since inception, with management of 2.1 million square feet of data center space.

Surging Investment Levels

The scale of capital flowing into data center infrastructure is unprecedented. The global compute power value chain is estimated to require $5.2 trillion in capital expenditure by 2030 to meet AI demand alone, with a range of $3.7 trillion to $7.9 trillion depending on adoption scenarios. On the demand side, leading hyperscalers are projected to invest more than $700 billion in 2026, approximately 60% year-over-year growth, fueled by the monetization of prior AI investments that is now showing up in accelerating cloud revenue. [25]

The Investment Spectrum

MEP (mechanical, electrical, plumbing) infrastructure comprises up to approximately 80% of a data center’s total value, with modern hyperscale developments approaching $1 billion per facility. Average development costs run approximately $11 million per megawatt, or roughly $1,500 per square foot, making data centers the most capital-intensive commercial property type to build. This capital intensity creates a natural investment spectrum ranging from powered land and powered shell to build-to-suit and value-add acquisitions, each with different risk/return profiles. Legacy enterprise-built facilities under 20 MW, many originally constructed at 2N redundancy, can be acquired below replacement cost and densified to generate value-add returns. [26]

Power Procurement as a Competitive Advantage

Dominion Energy, the primary utility serving Northern Virginia, projects approximately 7 GW of data center demand by 2030, growing at roughly 10% per year. New grid connections now take four to five years in most cases, and utilities are increasingly prioritizing applicants with development track records and upfront capital deposits over a first-come, first-served approach. In Virginia, data centers consumed 25% of the state’s electricity in 2025, a figure that could reach 46% by 2030. [27]

For experienced operators, these dynamics create a meaningful competitive moat. The ability to navigate utility queuing systems, secure power allocations, and maintain productive relationships with local utilities and regulators is not something new entrants can replicate quickly. In a market where power is the primary constraint, the entities that can deliver energized capacity fastest will capture disproportionate value.

Nuclear and Direct Power Procurement

Direct nuclear power procurement has emerged as a viable generation strategy for hyperscalers seeking reliable, carbon-free baseload power at scale. Over the past year, major technology companies collectively contracted for more than 10 gigawatts of new speculative nuclear capacity:

• Microsoft / Constellation Energy: 20-year PPA to restart Three Mile Island’s Unit 1 reactor (now the Crane Clean Energy Center). Timeline accelerated to 2027; supported by a $1 billion DOE loan. If completed, it will be the first reactivation of a fully shut U.S. nuclear plant.
• Amazon / Talen Energy: 17-year PPA for 1.92 GW from the Susquehanna nuclear plant, double the original commitment. Amazon has invested $20 billion+ to convert the adjacent site into an AI-ready campus and is also investing in X-energy small modular reactors.
• Meta / Constellation: 20-year PPA for 1.1 GW from the Clinton Clean Energy Center in Illinois, beginning 2027.
  Google / Kairos Power: First U.S. corporate SMR fleet deal for approximately 500 MW, targeting early 2030s delivery.
• Oracle: Plans for a 1 GW data center campus backed by three SMRs

[29]

The Case for Strategic Partnerships

The convergence of power constraints, regulatory complexity, and capital intensity means that aligning with experienced partners provides access to utility relationships, end-user relationships, local governance connections, and market intelligence that new entrants cannot replicate quickly. In a sector where the distance between a site and an energized substation can make or break a deal, operational experience is not a commodity; it is a competitive advantage.

Understanding the Investment Spectrum

Data center investment is not one-size-fits-all. The asset class offers a range of entry points across the risk-return spectrum, each with distinct capital requirements, execution complexity, and return profiles. Lincoln’s platform spans the full spectrum:

• Powered Land: Acquire raw or covered land at a downside-protected basis (e.g., industrial alternative use value). Target power procurement and data center entitlements within 12-36 months. Once a power agreement is secured, sell or recapitalize at a 2.0x+ multiple. This strategy focuses on large power requirements (100-500+ MW) and rewards speed of execution on utility engagement and entitlement.
• Vertical Development: Once a power agreement is secured and horizontal improvements are complete, pursue powered shell or turnkey development, ideally under a build-to-suit structure for a single end user. Development yield targets typically range from 7.5-10.0%+ yield-on-cost under open-book structures, with IRRs in the mid-to-high 20s+%.
• Value-Add / Core Plus: Acquire existing data centers with actionable upside through power expansion, densification, or new leasing. Focus on underutilized sites with near-term access to additional power or assets with below-market rents and mark-to-market opportunity. Target high-teens (%) IRRs.
• Core / Stabilized: Fully leased, operational data centers with creditworthy tenancy (typically investment-grade) under long-term leases. Stable yield for long-term hold, typically targeting low-teens (%) IRRs.

[30]

The Playbook: How Lincoln Executes

Lincoln’s approach focuses on identifying and executing risk-adjusted investments with tactical upside through four complementary strategies:
[31]:

Target raw and covered land opportunities across both major and emerging U.S. data center markets. Focus on satisfying growing hyperscale requirements (50-100+ MW). Initiate power procurement with local utilities immediately upon acquisition, targeting formal power confirmation within the underwritten hold period. Realize value through land sale to a developer or contribution to a vertical development venture.

Power Delivery and Land Value Creation

Successful delivery of scalable power is the single greatest driver of land value in data center development. The ability to engage utilities, navigate interconnection queues, and secure formal capacity allocations is the most valuable competency in data center development today. Developers who can compress the timeline from raw land to energized, entitled site create outsized value at each stage of the power procurement process.

Lincoln’s Competitive Advantage

Lincoln’s advantage in the data center market is the combination of deep sector expertise and national-scale execution capability:

• 30+ year track record in the data center business, with a $1.5 billion+ investment portfolio across 21 data center investments since inception and management of 2.1 million square feet of data center space.
• Longstanding relationships across data center, telecom, digital infrastructure, and technology industries, including direct access to decision makers at hyperscale/cloud providers, colocation operators, and Fortune 500 enterprise end users.
• National reach with local execution: Investment leadership, sourcing, and execution capabilities in every major U.S. market, with locally based teams that have developed trusted relationships in each community Lincoln serves.
• Deep expertise in land use, master planning, and entitlements, with a history of creating value for public-sector stakeholders and driving positive outcomes in complex regulatory environments.
• Alignment with blue-chip investment partners and global financial firms, providing access to institutional capital and sophisticated deal structuring.

[32]

AI’s Continued Impact

Data center power demand was remarkably flat from 2015 to 2019 despite workloads nearly tripling during that period, as efficiency improvements kept pace with growth. That dynamic has fundamentally reversed. Since 2020, efficiency gains have slowed just as AI-driven demand has accelerated, creating the steepest growth trajectory the industry has ever experienced. Data center power demand is now projected to increase 165-175% by 2030 versus 2023 levels. [34]

AI capacity as a share of total data center demand is anticipated to grow at 33% or more per year between 2023 and 2030, driven by both training workloads and the rapid proliferation of inference applications. Supporting analysis projects 3.5x growth in AI-specific capacity from 2025 to 2030 alone, with 125 incremental gigawatts of AI-related capacity needed over that period. [35]

Nuclear and Alternative Energy

As mentioned in the “Data Center as an Investment” section, direct nuclear power procurement has shifted from experimental to mainstream over the past year. The key developments (the Three Mile Island restart, the expanded AWS/Talen relationship, Meta’s and Google’s major commitments, and favorable regulatory signals from the current administration) all point toward nuclear playing a meaningful role in the next generation of data center power supply. The 462 MW NuScale small modular reactor design received Standard Design Approval from the NRC in May 2025, two months ahead of schedule, and executive orders have directed the Department of Energy to accelerate new nuclear development.[36]

Regulatory and Sustainability Pressures

The regulatory environment for data centers is evolving rapidly. Multiple states have introduced legislation in 2025-2026 conditioning new development on environmental assessments, grid impact studies, or temporary moratoriums. At the federal level, FERC has proposed new rules requiring large-load customers to bear the full cost of grid upgrades, and a coalition of 230+ environmental organizations has called on Congress to pause new data center construction until federal frameworks are established. [37]

For sophisticated operators and investors, this evolving landscape reinforces the importance of proactive community engagement, sustainable design, and strong utility relationships. Green certifications and environmental performance are increasingly functioning as prerequisites for obtaining permits and incentives in many jurisdictions, not just reputational differentiators. The firms best positioned to navigate this complexity are those with deep local knowledge, established stakeholder relationships, and a demonstrated track record of responsible development.