**South Korea’s AI Infrastructure Expansion: SK Telecom Targets 15GW Data Centre Capacity**
South Korea is accelerating its positioning as a global AI hub through a sweeping infrastructure expansion that encompasses data centres, semiconductors, and robotics. At the forefront of this initiative is SK Telecom (SKT), which has outlined plans to build up to 15GW of AI data centre capacity as part of the country’s broader “mega projects” strategy. These efforts are being developed in collaboration with major conglomerates like Samsung Electronics, SK Group, and government ministries, with initial investments totalling hundreds of trillions of won.
**A Multi-Billion Won Investment Across AI Infrastructure**
The South Korean government, alongside private sector leaders, has committed significant funding to develop the foundational elements of AI infrastructure. Samsung and SK Hynix are driving semiconductor production, while SK, GS Group, Naver, and SK Telecom are leading AI data centre development. According to Reuters, the initial phase involves approximately 550 trillion won in investment to construct 8.4GW of AI data centre capacity, with construction set to begin in the first half of 2028.
SK Telecom’s own 15GW target extends beyond this initial phase, representing a long-term vision to establish South Korea as a critical node in the global AI ecosystem. The company emphasizes that AI data centre development is dependent on advanced semiconductors, reliable energy solutions, and robust data centre construction and operations.
**Power, Cooling, and Global Demand Challenges**
In a regulatory filing, SKT noted that a single 1GW-class AI data centre could require around KRW 70 trillion, underscoring the immense financial and logistical challenges involved. The company is exploring funding models that include strategic partnerships with global technology firms and overseas capital.
SKT also highlighted a growing global data centre capacity shortage, citing McKinsey & Company forecasts that predict demand will grow by 19% to 22% annually, with supply struggling to keep pace. By 2030, global data centre demand could reach 171GW to 219GW, with the U.S. alone facing a potential shortfall of 15GW.
Domestically, South Korea is well-positioned to address these challenges, with strengths in high-bandwidth memory production, nuclear and LNG-based power supply, and extensive experience operating large-scale semiconductor facilities.
**Supporting the Ecosystem: Chip Investments and Robotics**
The AI infrastructure push is being reinforced by major chip investment programmes. Samsung plans to invest 400 trillion won in new semiconductor fabs in Gwangju and 56 trillion won in advanced high-bandwidth memory facilities in Cheonan and Onyang. Meanwhile, SK Group has outlined long-term plans worth approximately 1,100 trillion won for semiconductor projects and 1,000 trillion won for AI data centres, with annual investments exceeding 100 trillion won expected over the next decade.
South Korea’s national AI strategy also includes physical AI and robotics, with goals to develop industry-specific robots, Korean physical AI foundation models, data factories, and to train 10,000 AI robotics specialists over five years.
**Looking Ahead**
As SKT develops its data centre clusters in Ulsan and Jeolla, with phased capacities reaching 5GW by 2029, the full 15G roadmap remains a long-term objective. Jung Jai-hun, President and CEO of SKT, stated that the project is aimed at “preemptively preparing the computing infrastructure that the global AI ecosystem needs,” positioning South Korea as a core AI infrastructure hub in Asia.
***
*Source: “South Korea to expand AI infrastructure base with data centre, semiconductor, robotics projects,” Reuters, via Cloud Computing News.*
