The South Korean government said on Thursday it plans to distribute more than 10,000 graphics processing units (GPUs) purchased from Nvidia Corp. to small and medium-sized enterprises, startups, academic institutions and national artificial intelligence (AI) projects starting next February, as part of its broader strategy to cultivate a strong domestic artificial intelligence ecosystem.
The South Korean government recently spent 1.4 trillion won (approximately $947.2 million) to purchase approximately 10,000 GPUs from Nvidia, which is the first phase of a plan to purchase 50,000 GPUs by 2030.
These GPUs will be configured into large-scale clusters to support high-speed computing for AI model training and inference.
Applications for artificial intelligence development projects from industry, academia and research institutions will be accepted online until January 28, with each project supporting up to 256 H200 GPUs or 128 B200 GPUs.
The South Korean government added that some 6,000 B200 GPUs scheduled to be delivered later will be allocated to a nationwide initiative to develop indigenous artificial intelligence basic models.
Previously, Nvidia CEO Huang Jensen announced at the end of October that the company would cooperate with the Korean government and major companies such as Samsung Electronics to deploy 260,000 GPUs in South Korea to build a large-scale artificial intelligence factory.
According to the plan, the South Korean government will use more than 50,000 GPUs to build a national sovereign AI platform, and Samsung Electronics, SK Group, and Hyundai Motor Group will each receive more than 50,000 GPUs. Naver will receive 60,000 yuan.