Samsung Electronics and SK Group are planning a sweeping investment program of as much as $650 billion to build a new semiconductor cluster in Gwangju, South Korea, as global competition for artificial intelligence infrastructure accelerates and demand for AI memory chips surges.
The plan, discussed on June 26, reflects a broader effort by South Korea’s largest conglomerates to secure leadership in AI semiconductors while responding to a rapid global buildout of data centers by major technology companies.
Industry estimates suggest Microsoft, Alphabet, Meta Platforms and Amazon could invest more than $1 trillioncombined in AI data centers by 2026, driving unprecedented demand for high-bandwidth memory and advanced chip packaging.
The global investment race has intensified competition among chipmakers. Micron Technology, the world’s third-largest memory chip maker, is expanding aggressively with support from the U.S. government. The company currently operates wafer capacity of about 300,000 units per month and plans to add roughly 100,000 more by 2030 through new fabs in Idaho, alongside additional facilities in New York, Taiwan through Powerchip Semiconductor Manufacturing Corp., and Japan.
In response, Samsung Electronics and SK Hynix are accelerating construction of their Yongin semiconductor clusters by seven and 12 years, respectively, while also advancing plans for a new cluster centered in southwestern South Korea.
Samsung Electronics is expected to invest between $210 billion and $280 billion, while SK Hynix is planning $280 billion to $370 billion in the new Gwangju-centered cluster. The companies are seeking government support to accelerate infrastructure development, particularly in power and water supply.
The South Korean government said it will provide streamlined, one-stop support for both Yongin and the new southwestern cluster as part of a broader regional development strategy.
Samsung Electronics plans to build six fabs in Yongin, while SK Hynix plans four, significantly expanding combined monthly production capacity. Both companies said they view AI infrastructure demand as still in its early stages and are positioning for long-term expansion.
Separately, SK Group, Samsung Group and Naver are preparing to invest about $290 billion to build five AI data centers of roughly 1 gigawatt each across regions including Jeolla, Chungcheong, Gyeongsang and Gangwon provinces. The initiative aims to reduce dependence on global hyperscalers and support domestic AI foundation model development.
SK Vice Chairman Yoo Young-sang, who leads the group’s AI strategy committee, said in California on June 26 that while international cooperation remains essential, technological self-reliance is critical for long-term competitiveness.
Samsung affiliates are also expanding investment across multiple industries tied to the AI supply chain.
Samsung Display plans to invest up to $100 billion over the next decade in its Asan and Cheonan campuses to expand OLED and next-generation display production for applications including mobility, smart glasses, appliances and robotics.
Samsung Electro-Mechanics is preparing a multibillion-dollar investment in a new AI semiconductor substrate facility in Sejong, while Samsung SDI is expanding battery production capacity in Cheonan and energy storage systems in Ulsan to meet rising demand from AI-driven infrastructure and mobility markets.
In Busan, Samsung Electro-Mechanics is also expanding production of multilayer ceramic capacitors and semiconductor substrates.
Samsung Biologics is expected to expand R&D and manufacturing capacity in Incheon as part of efforts to position biopharmaceuticals as a “second semiconductor” growth engine.
Hyundai Motor Group is preparing major investment in Jeonbuk’s Saemangeum area to build a physical AI and future mobility hub. LG Group affiliates are also considering expanded AI-related semiconductor investments, including at LG Innotek’s Gumi facility, as part of broader collaboration with Nvidia on AI infrastructure development.
