Nvidia bets on South Korea in AI infrastructure race

Nvidia bets on South Korea in AI infrastructure race
Nvidia deepens South Korea AI ties

​Nvidia has struck a string of agreements with South Korean technology groups, tightening its links with a country that sits at the center of the global AI supply chain. The deals give Nvidia a clearer path to advanced memory chips and new data-center customers as demand for AI computing continues to strain the semiconductor industry.

Highlights

  • Nvidia signed South Korean deals with SK Hynix, SK Telecom, Naver, and Doosan.
  • SK Hynix will develop advanced memory for Nvidia’s AI infrastructure roadmap.
  • SK Telecom plans a gigawatt-scale AI cloud, with the first data center due in 2027.

SK Hynix secures a central role

According to Reuters, Nvidia announced partnerships with SK Hynix, SK Telecom, Naver, and Doosan during Chief Executive Jensen Huang’s high-profile visit to South Korea. The companies did not disclose the value of the agreements, but the most closely watched deal is with SK Hynix, one of the world’s largest memory-chip makers and a key supplier of high-bandwidth memory used in AI servers.

SK Hynix signed a multiyear technology partnership with Nvidia to develop next-generation memory for global AI data centers. The agreement is meant to help memory supply keep pace with Nvidia’s expanding roadmap, which now stretches beyond data centers into robotics, personal computers, and AI supercomputers.

Huang said SK Hynix has been Nvidia’s largest memory partner and would remain so. He also said Nvidia already buys “billions and billions of dollars” from the company each year and expects that to grow substantially. For SK Hynix, the agreement strengthens its position against Samsung Electronics and Micron Technology in the race to supply advanced memory for AI systems.

AI data centers move up the agenda

SK Telecom said it would build a gigawatt-scale AI cloud in South Korea using Nvidia technology, with the first AI data center expected to come online in 2027. Nvidia also said Naver and Doosan would use its technology to help build AI data centers.

Doosan, which develops robots and makes materials used in Nvidia’s Blackwell chips, expects its energy solutions to be used in Nvidia data-center platforms. Nvidia is also working with LG Group on electronics, mechanical systems, and AI for humanoid robots, as well as future data-center architecture, including cooling and power delivery.

The announcements landed during a sharp pullback in South Korean tech stocks. The Kospi had doubled over six months as SK Hynix and Samsung benefited from the AI boom, but it fell sharply on Monday as global technology shares came under pressure. SK Hynix and Samsung both opened more than 10% lower before trimming losses.

The supply-chain stakes

AI data centers need vast amounts of advanced memory, power, cooling, and network capacity. South Korea offers much of that in one place: SK Hynix and Samsung dominate memory production, while groups such as SK Telecom, Naver, Doosan, and LG can support data-center buildout, robotics, and industrial AI.

For Nvidia, securing memory supply is as important as selling chips. If demand for AI servers keeps rising, shortages in high-bandwidth memory could slow deployments and lift costs. 

We also reported Nvidia approves the Vera Rubin HBM4 supply from three major chipmakers. 

This material may contain third-party opinions, none of the data and information on this webpage constitutes investment advice according to our Disclaimer. While we adhere to strict Editorial Integrity, this post may contain references to products from our partners.
Weekly Top Bonuses
up to $2,500
deposit bonus for all clients
CLAIM BONUS
Your capital is at risk.