South Korea and Nvidia Partner too Substantially Expand National AI Infrastructure
A major collaboration between nvidia and the South Korean government is underway to dramatically expand the nation’s artificial intelligence infrastructure. The initiative will deploy over 250,000 Nvidia GPUs across both sovereign clouds and dedicated AI factories, aiming to fuel economic growth and innovation across key Korean industries like automotive, manufacturing, and telecommunications.
The Ministry of Science and ICT (MSIT) is spearheading the effort, initially focusing on deploying up to 50,000 of Nvidia’s latest GPUs to accelerate sovereign AI growth for businesses and industries nationwide. This expansion will begin with an initial deployment of 13,000 NVIDIA Blackwell and other GPUs, facilitated by Nvidia Cloud Partners Naver Cloud, NHN Cloud, and Kakao Corp. These partners will bolster computing infrastructure on Korea’s sovereign clouds, including through the establishment of the Korea’s National AI Computing Center.
The infrastructure will be accessible to research institutions,startups,and AI companies,providing resources to build and deploy AI models and applications,directly supporting Korea’s national strategy to enhance its AI capabilities.
Beyond cloud infrastructure, Nvidia is collaborating with Korean industry leaders – Samsung, SK Telecom, ETRI, KT, LGU+, and Yonsei University – on advancements in AI-RAN and 6G infrastructure. This collaboration aims to develop intelligent, low-power AI-RAN network technology that can reduce computing costs and improve device battery life by shifting GPU computation tasks to base stations.
Naver Cloud is independently expanding its Nvidia AI infrastructure, planning to deploy over 60,000 GPUs for both sovereign and physical AI applications. The company intends to leverage Nvidia Nemotron open models on its Nvidia AI infrastructure to develop specialized AI models for sectors like shipbuilding and security, with a focus on inclusive AI solutions for Korean citizens.
MSIT is also working with Nvidia on the Sovereign AI foundation Models project, utilizing Nvidia NeMo and open Nvidia nemotron datasets to develop sovereign language models trained on local data. Participants in this project include LG AI Research, Naver Cloud, NC AI, SK Telecom, and Upstage, enabling contributions from enterprises, researchers, and startups to create AI agents with advanced capabilities.
LG is further collaborating with Nvidia to promote the development of physical AI technology and build a robust physical AI ecosystem.
the Korea Institute of Science and Technology Facts (KISTI) is partnering with Nvidia to establish a Center of Excellence. This center will foster research collaboration utilizing Korea’s “Hangang” supercomputer – a sixth-generation national supercomputer powered by Nvidia accelerated computing. KISTI will also focus on building foundation models for scientific research and development, and supporting researchers in developing physics-informed AI models using the NVIDIA physicsnemo framework.
