Nvidia has announced plans to deploy up to 260,000 of its latest Blackwell GPUs in South Korea, partnering with the government and companies like Samsung Electronics, SK Group, Hyundai Motor Group, and Naver Cloud to build AI factories. The initiative will boost the country's AI computing capacity and offer significant opportunities for Samsung and SK hynix in supplying high-bandwidth memory chips. Nvidia CEO Jensen Huang praised South Korea's world-class memory technology during the APEC CEO Summit in Gyeongju.
At the APEC CEO Summit in Gyeongju on October 31, Nvidia announced plans to deploy up to 260,000 graphics processing units (GPUs) in South Korea to build large-scale artificial intelligence factories in partnership with the government and major companies. The initiative, previewed in a media briefing the previous day, will see CEO Jensen Huang share further details on the sidelines of the summit.
Allocations include 50,000 GPUs for the South Korean government to create a national 'sovereign AI' platform for training large language models optimized for Korean language and industrial applications. Samsung Electronics, SK Group, and Hyundai Motor Group will each deploy 50,000 units, while Naver Cloud installs around 60,000, boosting the country's total AI GPU capacity from about 65,000 to more than 300,000 and positioning it as a major global hub outside the U.S.
Huang arrived in South Korea on Thursday, meeting Hyundai Motor Group Executive Chair Euisun Chung and Samsung Electronics Chairman Lee Jae-yong in Seoul at the invitation of President Lee Jae Myung. He also discussed AI cooperation with SK Group Chairman Chey Tae-won.
The deployment is expected to benefit Samsung and SK hynix significantly, as each Blackwell GPU requires eight HBM3E memory units, totaling roughly 2.08 million for 260,000 GPUs—demand largely to be met by the Korean firms, excluding U.S.-based Micron. At an estimated US$300 per HBM3E chip, the value exceeds 900 billion won ($650 million). An industry official noted, 'The GPUs will be brought into South Korea, but the high-end memory chips that power them will come from Samsung and SK hynix.'
SK hynix leads with 64% of the global HBM market in Q2 sales, followed by Micron at 21% and Samsung at 15%. Samsung, having trailed recently, has shipped HBM4 samples to Nvidia, and the project could help balance competition. Huang stated, 'I am 100 percent confident that Samsung and SK hynix will be long-term partners... Korea is world class in memory technology.'
Additional partnerships include LG Group in robotics and medical technology, using its EXAONE model for AI-driven cancer diagnosis research. Nvidia will collaborate with Samsung, SK Telecom, KT Corp., LG Uplus, and Yonsei University on AI-native 6G networks. Hyundai plans to use its 50,000 GPUs for autonomous driving and smart factories, co-investing around US$3 billion with the government for a 'physical AI' ecosystem.