“SK Group and Nvidia will announce a business partnership between the two companies today… the size amounts to $500 billion,” said Nvidia CEO Jensen Huang during a meeting with South Korea’s President Lee Jae Myung in San Francisco. This enormous commitment marks one of the largest deals in the AI infrastructure sector, signaling a rapid acceleration of the race for AI dominance.
The collaboration centers on SK Group’s control over SK Hynix, a leading producer of high-bandwidth memory (HBM). HBM is key for AI systems, providing the ultra-fast stacked DRAM needed to support intensive data processing tasks in large AI clusters. With memory now becoming the main bottleneck in AI hardware development, SK Group’s role shifts from a typical chip supplier to a critical memory partner in Nvidia’s AI roadmap.
Recent reports mention SK Hynix postponing the launch of its next-gen HBM4 products to focus more on profitable commodity DRAM segments, while competitors like Samsung are advancing their own HBM4 technologies. This makes Nvidia’s partnership with SK Group strategically vital as global spending on AI infrastructure surges and players aim to secure reliable memory supplies for future systems.
This deal is part of a broader wave of South Korean initiatives involving giants such as Samsung, Naver, Hyundai, OpenAI, and Broadcom, all pushing to strengthen AI capabilities. The focus extends beyond chip speed to entire supply chains encompassing data centers and engineering talent. Nvidia’s recent moves follow its broader partnerships, including the $500 billion AI data center project announced with SK Group, reflecting a deliberate strategy to build out AI infrastructure on multiple fronts.
