The escalating demand for artificial intelligence (AI) is prompting a worldwide surge in data center development, a trend not observed since the peak of the industrial revolution. As AI applications proliferate, the need for computational power is reaching unprecedented levels. A recent analysis by Goldman Sachs suggests that the global appetite for AI-driven computing power could increase by 165% by 2030. Concurrently, McKinsey estimates that European markets will need between $250 billion and $300 billion in investments to meet the burgeoning information technology (IT) load, excluding necessary power generation capacity.
This growing demand is fundamentally reshaping the landscape of infrastructure investment, particularly in data centers—facilities crucial for cloud computing, application delivery, and the intensive training of AI systems. Data centers are now viewed not merely as technological assets but as vital components of modern economies, directly tied to advancements in AI and cloud services.
At present, the United States is leading this charge, hosting the most data centers globally. As of March 2023, there were approximately 5,426 data centers in the U.S., far surpassing Germany’s 529, the United Kingdom’s 523, and China’s 449. Projections indicate that these figures could expand by another 30-40% by the end of the decade, contributing to a global investment expectation of around $7 trillion in data infrastructure.
Land availability and energy requirements are becoming critical factors in the establishment of new data centers. According to Tim McGuire, Senior Director of Project Finance at Rowan Digital Infrastructure, land that can support data center development is becoming significantly more valuable. In high-demand areas such as Northern Virginia, the price of suitable land can exceed $2.5 million per acre. For hyperscaler companies—like Amazon Web Services, Google, and Microsoft—developers often look to acquire extensive parcels exceeding 100 acres to meet operational needs.
Energy and water are significant cost components in this regard, with energy availability posing particular challenges in various regions. McGuire explains that even in areas where energy infrastructure exists, creating the necessary interconnections can take months or years, with high associated costs. Consequently, utilities in regions experiencing substantial demand from data centers are increasingly requiring security measures before they can proceed with interconnects.
Jared Walter, a power and utilities analyst at EY-Parthenon, highlights disparities in energy access across regions. “Europe faces unique challenges regarding power availability, especially due to regulatory hurdles that complicate infrastructure expansion,” he states. “North America shares some of these difficulties, while Asia tends to facilitate quicker infrastructure development.”
Critical to AI’s functionality, graphic processing units (GPUs) are also focus areas in the context of data center development. Countries with fewer restrictions on GPU imports, such as Canada, Japan, and several European nations, are witnessing accelerated growth in their data centers, further fueled by advantageous local regulations. In contrast, some regions impose limitations that slow development progress.
The financing landscape for data centers is characterized by high capital intensity. Investors are typically large, globally recognized firms such as Blackstone. Claus Hertel, Managing Director at Rabobank, observes that relationships with these significant backers enable project financing that supports construction and operational phases. After a data center is operational, financing can shift towards permanent capital acquisition options, which may include asset-backed and mortgage-backed securities.
Hertel elaborates, “Once the center is commissioned, there is typically a window of three to four years where sponsors can secure long-term financing.” This transition reflects a broader trend whereby institutions prioritize securing favorable financing conditions ahead of finalized leasing agreements.
Rowan Digital Infrastructure operates within this paradigm as well. McGuire describes their approach: “Private equity investors often cover early-stage costs like land acquisition. We delay debt financing until we acquire signed leases, allowing access to highly competitive debt terms, given the strong financial ratings of our hyperscaler customers.”
The outlook for data center investment remains robust, further emphasizing a growing alignment between demand drivers and capital availability. Gordon Bell, a principal at EY-Parthenon, notes, “The industry has expanded dramatically in recent years, and that momentum should continue. Legacy asset owners are now seeking diverse strategies for raising capital to cater to escalating demands.”
Emerging financial mechanisms—such as dedicated funds that aggregate a portfolio of stabilized data center assets—are establishing pathways for a broader range of investors. Bell suggests that such funds would allow investors to mitigate risk while exposing them to a diversified array of markets and customer segments.
In contrast to the historical trend where investments were often concentrated in individual facilities, the formation of leveraged investment vehicles that encompass multiple facilities could enable scalable growth opportunities, attracting a wider pool of investors.
Despite the ongoing challenges associated with financing and infrastructure development, the consensus among industry experts is largely optimistic. As hyperscalers and revenue generation in cloud services continue to flourish, the demand for data centers is expected to remain strong. McGuire concludes, “AI will undoubtedly be a driving force, but the appetite for cloud services will play an equally critical role in the future of data center investments.”
In conclusion, the significant investments in data centers underscore their integral role in supporting the burgeoning AI landscape while also exemplifying the intricate interplay between computing power, capital investment, and regulatory environments shaping the global economy. As stakeholders in this industry strategize for the future, it is apparent that the intersection of technology and finance will not only dictate the pace of development but also redefine the parameters of innovation and competitiveness in an ever-evolving economic landscape.