The South Korean government plans to begin distributing approximately 10,000 advanced graphics processing units (GPUs) to industry, academia, research institutions, and national AI projects starting in February next year, to boost the domestic AI ecosystem. This initiative is part of the government’s ‘AI expressway’ project, aiming to secure over 52,000 GPUs by 2028.
The Ministry of Science and ICT selected Naver Cloud, Kakao, and NHN Cloud as project participants through a public bidding and evaluation process in July. These companies are tasked with securing and building a total of 13,000 GPUs, including NVIDIA B200 (10,080 units) and H200 (3,056 units). The government will utilize 8,160 B200 units and 2,296 H200 units among them.
Each participating company will complete GPU deployment through cooling system installation, large-scale clustering, optimization, and beta testing. A pre-beta test is currently underway for 512 B200 units to verify procedures and identify potential user inconveniences before full government deployment.
To ensure fair and efficient allocation, the government has established a distribution framework prioritizing national AI projects and initiatives in industry, academia, and research. Next year, the government plans to continue securing GPUs through the construction of Supercomputer 6 (9,000 units) and additional purchases (15,000 units), with the goal of launching the National AI Computing Center by 2028.