韩国数据中心投资：稀缺溢价、供应与退出潜力

Diagnosing Scarcity Premiums, Supply & Exit Potential INTRODUCTION Korea’s data center market has experienced an unprecedented boom over the past fiveyears, with domestic and international asset managers and global private equity fundsundertaking large-scale projects in Greater Seoul. With several major transactionshaving been completed and a robust deal pipeline forming, data centers are now firmlyestablished as a core alternative asset class. At the root of this growth is the structural expansion of computing demand driven bygenerative AI. Global big tech infrastructure investment is setting new benchmarkseach quarter, providing strong support to data center demand, with Korea among thebeneficiaries of this global expansion. In 2026, however, the market is entering a critical inflection point. Secured land andgrid power availability in Greater Seoul have reached practical limits, triggering astructural supply cliff, while the sustainability of AI workload demand and exitstructures are being questioned. This report cross-analyzes global demand structures with Korea's supply constraints;outlines entry requirements by potential buyer type; and charts capital circulationpathways. It concludes by providing solutions for developers, institutional investors,and capital providers seeking viable exit routes. Contents 01 Global AI Trends and Analysis of Advanced U.S. Markets 02 Korea Data Center Market Analysis 03 Korea Data Center Investment Market Analysis 04 Outlook and Recommendations Global AI Trends andAnalysis of Advanced U.S. Markets01 AI Transition Triggers Structural Surge in Power Demand, Diversifying Global Beneficiary Regions Driven by the launch of ChatGPT in late 2022, recent years have seen theglobal data center market undergo a structural paradigm shift towardhigh-performance computing infrastructure. The proliferation ofgenerative AI has triggered hyperscale demand for Large LanguageModel (LLM) training and inference, while the rise in AI workloads—unfolding much faster than the cloud transition era—is drivingexponential growth in global power demand. Figure 1 shows that global data center power consumption is projected tosurge more than 5x from approximately 200 TWh in 2015 to more than1,100 TWh by 2030. The share of AI within total workloads is estimated toexpand steeply from less than 3% prior to 2023 to 13% in 2025 and 29% by2030, indicating that a significant portion of future power growth will beallocated to AI operations. While the U.S. currently anchors global power demand, future AI-driveninfrastructure demand is set to expand worldwide rather than beingconfined to specific geographical regions. With synchronized growthbecoming visible globally, including in Asia, the steep growth trajectory ofthe Rest of World AI segment within the chart is expected to beparticularly prominent. This decentralized global demand trend is projected to serve as a newopportunity for major Asia Pacific hub markets. As the AI workloaddemand curve outside the U.S. slopes steeply upward, the focus of globalbig tech companies seeking to expand infrastructure and secure new datacenters will gradually diversify beyond the U.S. into major Asia Pacific hubmarkets, such as Korea. Global Tech Big FourCapExSignals Structural, Cycle-Independent Data Center Demand The acceleration of the global generative AI market is directly translatinginto a surge in CapEx by big tech companies, structurally supporting datacenter demand. The combinedCapExof the four largest global techcompanies-Microsoft, Meta, Google, and Amazon—has shown a steepincrease since 2022 when generative AI took off, serving as the strongestfundamental driving the growth of the data center market. Bloomberg and market research data show these firms’ combinedCapEx,which stood at approximately USD 200 billion in 2024, surged to aroundUSD 330 billion in 2025 and USD 610 billion in 2026. It is estimated thatthe big four tech companies will deployCapExranging from a minimum ofUSD 800 billion to a maximum of USD 1.1 trillion in 2027. This surge in capital spending is not a simple cyclical investmentexpansion, but rather a structural increase in expenditure aimed atadvancing generative AI services and expanding hyperscaleinfrastructure. With major global demand drivers publicly expanding theirmid-to-long-termCapEx, the sustainability of future AI demand appearssolid. Competition among major big tech companies to preemptively securedata centers is therefore expected to remain intense. Coupled with thegrowth in global demand mentioned in the previous chapter, these firms’massive capital deployment and commitment to securing infrastructurewill likely act as a key driver that further intensifies pressure forpreemptive infrastructure expansion into Asia Pacific and Korea. AI Penetration Gap Between Theory and Reality—Evidence of Early-Stage Demand Capture Despite significant capital deployment by big tech companies, the actualpenetration of AI