U.S. chipmaker Nvidia announced on Monday a major data center project in South Korea with SK Telecom as part of several business agreements in the country. The world’s most valuable company also revealed plans to collaborate with memory chip producer SK hynix on advanced components essential for AI systems that remain in short supply. The announcements followed chief executive Jensen Huang’s weekend meetings with South Korean technology executives in Seoul. SK Telecom and Nvidia intend to construct a gigawatt-scale AI cloud facility in Korea, with the first AI factory expected to begin operations in 2027. The project aims to provide sovereign, physical and agentic AI services for local businesses and industries, with potential expansion across Asia. No investment amount was disclosed. SK hynix, under the same parent group as SK Telecom, separately confirmed a multi-year technology partnership with Nvidia focused on memory chips. The deal seeks to ensure supply for advanced memory amid lengthy development timelines and high capital needs for global AI infrastructure expansion. SK hynix will also enter new markets created by Nvidia, including AI infrastructure, personal AI and physical AI, by co-developing memory components. As governments and firms invest hundreds of billions in AI infrastructure, Nvidia’s market value has exceeded $5 trillion. The push to build AI data centers has triggered a worldwide shortage of memory chips, boosting profits for producers such as SK hynix and Samsung Electronics. SK Group chair Chey Tae-won recently pledged to double silicon wafer production capacity but warned shortages may last until 2030 due to long factory construction times. Nvidia also revealed AI collaborations with Naver and Doosan Group on robotics. The company is primarily known for GPUs that power AI applications from chatbots to image generators. Last week Nvidia introduced a high-performance laptop chip for Windows devices to compete in the AI-integrated consumer PC market.
