Nvidia and South Korea’s SK Group have unveiled an artificial intelligence infrastructure initiative valued at more than $500 billion, covering large-scale AI data centres and the development of next-generation memory technologies.
The initiative includes a long-term partnership between Nvidia and SK Hynix aimed at securing supplies of advanced memory for future AI computing systems. The companies will also work together to develop high-bandwidth memory for AI model training, AI agents and physical-AI applications such as robotics and autonomous systems.
As part of the initiative, SK Telecom plans to develop a two-gigawatt AI data centre powered by Nvidia’s upcoming Vera Rubin computing platform and SK Hynix’s HBM4 high-bandwidth memory. The first facility under the project is expected to become operational in 2027.
The collaboration brings together Nvidia’s AI computing platforms, SK Hynix’s advanced memory capabilities and SK Telecom’s data-centre infrastructure. It comes as demand for AI computing capacity and high-bandwidth memory continues to rise globally, encouraging technology and semiconductor companies to secure long-term supply arrangements.
Separately, Nvidia, South Korean internet company Naver and infrastructure investment firm Brookfield plan to expand Naver’s AI data-centre capacity in South Korea. Financial and capacity details of the proposed expansion were not disclosed.
The initiative could also have implications for India’s expanding AI infrastructure market. Greater production and development of HBM and advanced AI platforms could support the global availability of computing systems required by Indian data-centre operators, cloud-service providers and enterprises. However, no India-specific investment or deployment has been announced as part of the initiative.
The announcement reinforces South Korea’s position in the global AI supply chain, particularly through its established capabilities in memory-chip manufacturing and telecommunications infrastructure.
