U.S.-based technology company Nvidia announced on Friday its plan to supply more than 260,000 of its next-generation Blackwell artificial intelligence chips to the South Korean government and several of the nation’s leading technology companies, including Samsung Electronics.
The agreement focuses on providing South Korea’s public and private sectors with advanced hardware for AI development and data processing. The announcement did not specify the exact models to be supplied, but the Blackwell architecture is the successor to Nvidia’s Hopper H100 and H200 chips, which are currently used in data centers worldwide for training and deploying large language models.
Nvidia unveiled the Blackwell platform in March 2024, with company officials stating at the time that the new chips offer substantial improvements in computing performance and energy efficiency for AI workloads. According to Nvidia, the technology is designed to support the development of AI models with trillions of parameters.
The procurement aligns with South Korea’s national strategy to strengthen its domestic AI capabilities. The South Korean government has publicly committed to significant investments in building a sovereign AI ecosystem, aiming to bolster its position in the global technology landscape. The acquisition of advanced semiconductors is a central component of this initiative, intended to provide the necessary computing power for research institutions and corporations.
Samsung Electronics, a major recipient mentioned in the announcement, is a global leader in memory chips and consumer electronics and has been expanding its focus on AI technology. Other major South Korean technology conglomerates, such as SK Hynix and Naver, are also key players in the country’s AI sector, though they were not explicitly named in Nvidia’s initial statement.
This supply agreement occurs amid high global demand for high-performance AI processors. Technology companies and governments are competing to secure the computational resources required for advancements in generative AI and other data-intensive applications. The distribution of these components is subject to international trade regulations, including export controls maintained by the U.S. government on certain high-end semiconductor technologies.
Details regarding the financial terms of the deal and a specific timeline for the delivery and deployment of the chips were not released.
