Nvidia announced during the APEC meetings in Gyeongju that it will supply South Korea with more than 260,000 advanced AI GPUs to address the nation’s shortage of computing power for training complex AI systems. The initiative involves collaboration with the Ministry of Science and ICT, Samsung, SK Group, Hyundai Motor Group, and cloud providers Naver and Kakao, aiming to expand domestic AI infrastructure through government and private sector investments valued by local reports at up to $10 billion over several years.
The agreement positions South Korea to significantly enhance its AI capabilities. Bae Kyung-hoon, the deputy prime minister and minister of Science and Information and Communication Technologies, stated, “This will support South Korea’s prosperity as it strives to become one of the top three global AI powerhouses.” This partnership addresses ongoing challenges in acquiring sufficient high-performance GPUs, which are essential for developing and refining sophisticated AI models that require extensive computational resources.
South Korea currently faces constraints in its AI development due to limited access to powerful GPUs. The incoming supply from Nvidia represents one of the largest such procurements by any country and will more than quadruple the nation’s existing AI chip inventory. Ha Jung-woo, presidential secretary for AI and future planning, provided details from a closed-door meeting in a statement to The Korea Herald: “In a closed-door meeting, Huang said the US has about 20 million GPUs, the most in the world, followed by China. Korea’s total of 300,000 now puts it in third place globally.” This comparison underscores the scale of the expansion, elevating South Korea’s position in global AI computing capacity.
The government component of the deal focuses on establishing controlled domestic infrastructure. Through the Ministry of Science and ICT, over 50,000 of Nvidia’s latest GPUs will be deployed across the forthcoming National AI Computing Center and partnerships with local cloud providers including NHN Cloud, Kakao, and NAVER Cloud. These resources will enable researchers, startups, and companies to develop AI models independently, reducing dependence on external technology platforms and fostering self-reliant innovation in AI applications.
Major South Korean conglomerates are also committing substantial resources to integrate the GPUs into their operations. Samsung plans to construct an “AI factory” featuring more than 50,000 GPUs, dedicated to advancing semiconductor development and creating sophisticated digital systems that support broader technological ecosystems. This facility will process vast datasets to optimize chip designs and simulate complex manufacturing processes, directly contributing to improvements in electronic components and related industries.
SK Group intends to launch Asia’s first industrial AI cloud, utilizing Nvidia’s RTX PRO 6000 Blackwell chips. This platform will facilitate AI-driven solutions for industrial applications, such as predictive maintenance in manufacturing and optimization of supply chains, by providing scalable computing power tailored to enterprise needs across various sectors. The initiative builds on SK’s expertise in chemicals, energy, and telecommunications to integrate AI into core business functions.
Hyundai Motor Group will allocate 50,000 GPUs specifically to its autonomous driving initiatives and smart manufacturing programs. These GPUs will power simulations for vehicle autonomy, including real-time data processing for sensor fusion and path planning algorithms, while also enhancing factory automation through AI models that predict equipment failures and streamline production lines for electric and conventional vehicles.
Naver Cloud is set to bolster its infrastructure by incorporating 60,000 additional GPUs, which will expand capacity for training larger AI models and delivering enhanced enterprise services. This upgrade supports Naver’s ecosystem of search, e-commerce, and content platforms by enabling faster inference times and more accurate personalization features, while also offering cloud-based AI tools to external businesses for custom model development.
Nvidia CEO Jensen Huang commended the collaborative effort, stating, “Just as Korea’s physical factories have inspired the world with sophisticated ships, cars, chips and electronics, the nation can now produce intelligence as a new export that will drive global transformation.” His remarks highlight the transition from traditional manufacturing strengths to AI as a strategic asset.
The arrangement occurs against the backdrop of U.S.-China trade restrictions that limit Nvidia’s sales of advanced chips to China, prompting increased engagement with other Asian markets like South Korea. During press interactions, Huang indicated interest in future sales of Blackwell-series chips to China but clarified that no such plans exist at present. Deployment of the GPUs will commence next year and extend throughout the decade, aligning with long-term AI infrastructure goals. This development further solidifies Nvidia’s leadership in the AI chip sector, following its achievement as the first company to reach a $5 trillion valuation.
