Nvidia will help build a network of large AI computing centers in South Korea by deploying up to 260,000 Blackwell GPUs with the government and major conglomerates, according to Yonhap News Agency. The plan would lift Korea’s installed AI GPU count from roughly 65,000 to more than 300,000, positioning the country among the largest AI-compute hubs outside the United States.
The Seoul government will allocate up to 50,000 GPUs to assist a national “sovereign AI” platform that is designed for Korean language and business use. According to Yonhap, Samsung Electronics, SK Group, and Hyundai Motor Group will each set up to 50,000 units. Naver Cloud will set up roughly 60,000.
On Friday, Nvidia CEO Jensen Huang is anticipated to give more information on the side of the APEC CEO Summit in Gyeongju. Huang got to Seoul on Thursday and met with Lee Jae-yong, the chairman of Samsung, and Euisun Chung, the executive chair of Hyundai. The financial details were not made public.
Companies sketched early use cases: Samsung aims to speed chip manufacturing with digital twins and smart robotics using Nvidia’s NeMo, CUDA-X and Omniverse platforms; SK Group plans semiconductor R&D and an industrial AI cloud; SK Telecom will offer sovereign industrial cloud services; Hyundai intends to train large models for autonomous driving, robotics and smart factories, with an estimated co-investment of about $3 billion with the government; and Naver Cloud will provide backbone compute for domestic foundation models, Yonhap said.
LG Group will partner with Nvidia in robotics and medical technology, including AI-driven cancer diagnosis research using LG AI Research’s EXAONE model. Nvidia also plans to work with Samsung, SK Telecom, KT, LG Uplus and Yonsei University on AI-native 6G radio access networks.
A global media for the latest news, entertainment, music fashion, and more.
