Nvidia and South Korea's SK Group have announced a partnership worth more than $500 billion aimed at building out the infrastructure behind the artificial intelligence boom.
According to Reuters, the two companies unveiled the initiative on Friday in San Francisco, describing it as spanning large-scale AI data centers and next-generation memory chips. The deal ties together two critical layers of the AI stack: the processors that run AI models and the high-speed memory that feeds them data.
At the center of the plan is a 2-gigawatt AI data center. The Economic Times reports that SK Telecom intends to build the facility using Nvidia's next-generation Vera Rubin chips paired with SK Hynix's HBM4 high-bandwidth memory, with the first site expected to come online. As part of the arrangement, Reuters notes that SK Hynix will help secure next-generation memory supply for Nvidia and jointly develop HBM chips.
Bloomberg, via Techmeme, reports the partnership extends beyond SK Group: Nvidia plans to invest $1 billion in Naver Corp. to help finance an AI data center already under construction in South Korea, while the SK collaboration targets more than 2 gigawatts of AI data center capacity.
The announcements came as Nvidia CEO Jensen Huang met with South Korean President Lee Jae Myung. The Korea Herald reports that Huang pledged massive investment and cooperation with South Korea during the meeting.
Why it matters: the deal underscores how tightly the AI race now depends on securing both cutting-edge chips and the specialized memory to run them—and how much money the world's largest tech players are willing to commit to lock up that supply.