The Booming Data Center Industry: A Look at Investor Requirements, Site Selection, and a Comparison to Puerto Rico's Pharmaceutical Industry

Introduction

As evidenced by a global spending spree, the rise of data centers as foundational infrastructure underscores their critical role in powering artificial intelligence, cloud computing, and digital economies. From an investor perspective, jurisdictions considering the possibility of a data center being located in their area must meet at least two conditions: energy at scale and fast regulatory approval.

The Wall Street Journal quotes Roy March, CEO of Eastdil Secured, on this trend: “New data centers are focused on markets with ‘the ability to deliver power in scale and jurisdictions that can fast-track it’” (Grant, 2025). Other investors have expressed similar observations (Note 1).?

Our article examines the importance of data centers and details their planned and under-construction locations across U.S. states and territories. It contrasts this boom with Puerto Rico’s pharmaceutical manufacturing surge, exploring why the island lacks viability as a data center hub despite its historical pharmaceutical success. This suggests that Puerto Rico is being bypassed by the current boom, in contrast to the “golden era” of the pharmaceutical manufacturing surge.

The Necessity of Data Centers

Data centers are indispensable to the digital era, enabling the storage, processing, and distribution of data that drives AI, cloud services, and real-time applications. The proliferation of generative AI (e.g., ChatGPT) and hyperscale computing has escalated demand, with U.S. data center power needs projected to hit 35 gigawatts (GW) by 2030 (McKinsey, 2024). Facilities now require 100–480 megawatts (MW) to power over 100,000 homes, as exemplified by Utah’s 175 MW campus (Grant, 2025). This boom, driven by a 1.5x data increase from 2022 to 2024 (KKR, 2024), supports critical sectors from e-commerce to scientific research, offering utility-like returns (3.14% yield, NAREIT, 2022).

March’s insight highlights two imperatives: power scalability and speed. AI workloads, such as xAI’s Colossus supercomputer in Memphis, demand gigawatt-scale electricity without grid constraints. At the same time, competitive tech markets require rapid deployment to circumvent 3-5-year interconnection delays (Swenson, cited in Grant, 2025). With digital infrastructure contributing 4%+ of global emissions (Investments and Wealth, 2024), data centers’ necessity reshapes energy systems and real estate markets, aligning with economic and technological mandates.

Locations of Data Centers in U.S. States and Territories

The data center boom spans a diverse array of U.S. states, prioritizing regions with robust power infrastructure and regulatory agility, as March emphasizes. Below is a table of planned, under-construction, and recently operational data centers based on current development data:

Key Trends:

• Fast-Track Leaders: States like Texas (ERCOT), Virginia (Loudoun’s program), and Tennessee (TVA) host 6.3 GW under construction in 2024 (CBRE), with investments from $285M (Illinois) to $10B (Louisiana, Ohio). Fast-track status hinges on streamlined permitting and utility cooperation.
• Power Diversity: Nuclear (Tennessee, Illinois), gas (Texas, Louisiana), and renewables (Oregon, Wyoming) ensure scalability. Operational since 2024, xAI’s Memphis facility leverages TVA’s nuclear/gas mix.
• Emerging Hubs: Alabama, Nebraska, and Wyoming join traditional leaders, driven by land and power availability, while Puerto Rico remains unlisted despite its potential.

Puerto Rico’s Position: With potential sites (JLL, 2021), Puerto Rico could align with fast-track goals via Act 60 (4% CIT, 75% property tax relief), but its grid—70% fossil fuel, 2% renewable—lacks the scale March deems critical, unlike states with gigawatt-ready infrastructure.

Comparison with Puerto Rico’s Pharmaceutical Manufacturing Boom

Puerto Rico’s pharma boom offers a historical parallel to assess data center feasibility, revealing why the island struggles to replicate its past success.

• Drivers and Incentives:

1. Pharma (1976-1990): Section 936’s 100% federal tax exemption (1976-2006) catalyzed a $44 billion export industry, drawing giants like Pfizer. Manufacturing hit 50% of the GDP, fueled by small-molecule drug demand and U.S. market access.
2. Data Centers (2025): AI and cloud growth drive a $1 trillion U.S. boom (Blackstone), with Act 60s 4% CIT and partial exemptions offering competitive but less potent incentives. According to March, power, not tax breaks, is the primary draw.

• Infrastructure:

1. Pharma: Plants (50,000-200,000 sq ft) used 10-50 MW, supported by a then-sufficient grid. Industrial hubs like Barceloneta thrived without gigawatt-scale needs.
2. Data Centers: Campuses (e.g., 100 acres, 175 MW in Utah) require 100+ MW near nuclear, gas, or renewables. Puerto Rico’s 2,578 MW capacity (EIA, 2023) is unreliable post-Maria and misaligned with March’s scalability standard.

• Investment Scale:

1. Pharma: Billions accrued over decades ($100M-$500M per plant), with steady growth until 2006. Local tax breaks ($100.5B, 2017-2023, CPD) sustained legacy operations.
2. Data Centers: Rapid investments ($1B-$10B per project) reflect a trillion-dollar surge, dwarfing pharma’s pace. Puerto Rico’s Act 60 lacks the tailored pull of states like Texas.

• Why It Is Not Feasible Now in Puerto Rico:

1. Power Scalability Deficit: March’s “ability to deliver power in scale” is unmet. With 70% oil/gas and 2% renewables, Puerto Rico's grid can’t match Tennessee’s nuclear (41%) or Oregon’s hydro (50% +). Post-Maria outages persist despite $20B in HUD funds.
2. Infrastructure and Speed Lag: Act 60s 30-90-day permitting aligns with fast-track goals, but grid upgrades and environmental hurdles (e.g., Clean Water Act) lag behind states, avoiding 3-5-year delays. March’s speed criterion is compromised.
3. Competitive Disadvantages: Pharma shifted to Asia (80% APIs, PhRMA) and U.S. hubs (e.g., Lilly’s $5.5B Indiana plant) post-Section 936. Data centers favor power-rich states (e.g., Virginia’s 13.4M sq ft), outpacing Puerto Rico’s tax-driven appeal.
4. Economic Context: A $70B debt crisis (2017) and higher labor costs ($15-$20/hour vs. $2-$5 in India) erode competitiveness, unlike the stable, tax-fueled environment of the 1970s.

Finally

Data centers are essential in the digital age, their necessity rooted in the computing demands of AI and our reliance on digital technology, as highlighted by their presence in 15 U.S. states with investments reaching $10 billion. March’s?emphasis on power scalability and expedited regulatory processes fuels this growth, from Utah’s 175 MW capacity to Nebraska’s 7,000 MW potential. Puerto Rico’s pharmaceutical boom, driven by tax incentives, depended on relatively modest infrastructure and unique advantages that cannot be easily replicated today.?The island’s grid fragility, competitive lag, and economic instability hinder its data center prospects, despite Act 60's promise. To echo its pharma past, Puerto Rico must modernize power—perhaps via renewables or microgrids—to meet March’s criteria and reclaim its investment stature in 2025.

References

• Grant, P. (2025, March 4). “Banks Loan $2 Billion to Build a 100-Acre AI Data Center in Utah.” Wall Street Journal.
• Invest Puerto Rico. (2021). Economic Data Reports.
• McKinsey & Company. (2024). Data Center Power Demand Outlook.
• KKR. (2024). Global Data Growth Trends.
• Blackstone. (2025). Infrastructure Investment Projections.

Notes:

1. Investors with Similar Observations

Larry Fink (CEO, BlackRock)

• Ability to Deliver Power in Scale: Larry Fink has highlighted the surging energy demands of AI and digital infrastructure as a transformative investment opportunity. In BlackRock’s 2025 outlook and prior statements, he notes that data centers’ power needs—projected to grow 5x- 7x in the U.S. over the next 3-5 years—require proximity to scalable energy sources like nuclear, natural gas, and renewables. For instance, BlackRock’s investment in the Waratah Super Battery in New South Wales focuses on supporting grids strained by tech growth, a principle applicable to U.S. markets like Virginia or Texas.
• Jurisdictions That Can Fast-Track It: Fink’s strategies emphasize regions with regulatory agility that can deploy infrastructure quickly. BlackRock’s infrastructure portfolio targets areas with streamlined permitting—like Texas or Tennessee—where utilities and governments expedite power projects to meet tech demand. This mirrors March’s focus on speed as a competitive edge.
• Example: Fink’s discussions at the BlackRock Future Forum underscore how infrastructure modernization, including rapid grid upgrades, drives economic and portfolio growth, aligning with March’s view on power and speed.

Hussain Sajwani (Chairman, DAMAC Properties)

• Ability to Deliver Power in Scale: On January 7, 2025, Sajwani, an Emirati billionaire, announced a $20 billion investment in U.S. data centers, citing AI and cryptocurrency growth as drivers. His focus on the U.S. reflects an understanding that gigawatt-scale power—available near nuclear plants (e.g., Tennessee) or renewable hubs (e.g., Texas wind)—is essential for these facilities, echoing March’s observation about the Utah data center’s 175 MW benchmark.
• Jurisdictions That Can Fast-Track It: Sajwani’s investment, tied to Trump-era policy optimism, targets states with fast-track potential like Texas and South Carolina, where tax incentives and quick permitting have historically lured tech projects. His move aligns with March’s emphasis on speed, as delays in grid interconnection could jeopardize such a massive capital commitment.
• Example: Per Data Center Frontier, Sajwani’s strategy leverages U.S. regions with robust power infrastructure and regulatory support, directly paralleling March’s criteria.

Sam Altman (CEO, OpenAI)

• Ability to Deliver Power in Scale: Altman’s ambitious Stargate Project, a $500 billion AI data center initiative announced January 21, 2025, hinges on securing vast energy resources—potentially terawatts—through partnerships with Oracle, Microsoft, and others. He’s advocated for nuclear power (e.g., SMRs) to meet AI’s constant demand, aligning with March’s focus on scalable, bottleneck-free energy near generation sources.
• Jurisdictions That Can Fast-Track It: Altman’s collaboration with U.S.-based tech giants suggests a preference for states like Nevada or Ohio, where regulatory frameworks and utility cooperation enable rapid deployment. This mirrors March’s point about avoiding 3-5-year delays, as Stargate’s timeline demands speed to maintain AI leadership.
• Example: Data Center Frontier notes Altman’s reliance on power-rich, fast-track regions, reinforcing March’s market reality of energy and speed as twin pillars.

Brian Venturo (Co-Founder, CoreWeave)

• Ability to Deliver Power in Scale: Venturo, whose company builds specialized AI data centers, has stressed the need for proximity to gigawatt-scale power sources like hydroelectric dams or gas plants. CoreWeave’s expansion into secondary markets (e.g., Indiana and Iowa) reflects a strategy to tap underutilized power infrastructure, similar to March’s Utah example, where 175 MW supports a single campus.
• Jurisdictions That Can Fast-Track It: Venturo targets states with accelerated permitting and utility responsiveness, such as Texas and Georgia, to deploy AI infrastructure quickly. Novva's Wes Swenson notes that the expedited process echoes March’s observation that competitive tech real estate hinges on avoiding delays.
• Example: Industry reports (e.g., Investments and Wealth) highlight CoreWeave’s focus on power scalability and regulatory speed, matching March’s investment logic.

Michael Callari (Business Development Manager, TPT Retirement Solutions)

• Ability to Deliver Power in Scale: While Callari’s role is pension-focused, his firm’s $6 billion investment in data centers via Harrison Street reflects a broader institutional recognition of power-intensive tech growth. These investments target regions like Virginia and Texas, where grid capacity supports hundreds of megawatts, aligning with March’s scalability benchmark.
• Jurisdictions That Can Fast-Track It: TPT’s strategy favors fast-track states with economic incentives (e.g., South Carolina, Nebraska), where rapid permitting and utility upgrades enable quick returns on infrastructure bets. This strategy parallels March’s emphasis on speed as a market differentiator.

Example: MarketsGroup ties Callari’s focus on data centers to power-rich, agile jurisdictions, reinforcing March’s dual criteria.

Final Remarks

A group of friends from “Organizational DNA Labs” (a private group) compiled references and notes from various theses, authors, and academics for the article and analysis. We also leveraged AI platforms such as Gemini, Storm, Grok, Open-Source ChatGPT, and Grammarly as research assistants to save time and check for expressions' structural and logical coherence. Our use of these platforms aims to verify information from multiple sources and validate it through academic databases and equity firm analysts with whom we have collaborated. The references and notes in this work provide a comprehensive list of the sources utilized. As the editor, I have taken great care to ensure that all sources are properly cited and that the authors are duly acknowledged for their contributions. The content is primarily based on our analysis and synthesis of the sources. The compilation, summaries, and inferences reflect our dedication and motivation to expand and share our knowledge. While we have drawn from quality sources to inform our perspective, the conclusion represents our views and understanding of the topics covered as they evolve through ongoing learning and literature reviews in this business field.