South Korea Secures 260,000 Nvidia Chips to Advance National AI Strategy

Nvidia (NVDA) announced on Friday that it will provide up to 260,000 artificial intelligence (AI) chips to South Korean corporations and the national government as part of the country’s extensive initiative to establish AI sovereignty.

The declaration, made during the APEC Summit in South Korea, includes multiple agreements across both public and private sectors. Among the participants are four of South Korea’s largest companies — Samsung, SK Group, Hyundai, and NAVER Cloud. According to Nvidia, these partnerships will increase the total number of its AI chips in use across South Korea from approximately 65,000 to more than 300,000. The company declined to disclose the financial details or specific timelines for delivery.

Nvidia emphasized that the initiative aligns with South Korea’s efforts to achieve AI sovereignty — a growing global movement in which nations seek to build domestic AI infrastructure and develop independent AI models. In June, Nvidia projected that the sovereign AI market could expand to $1.5 trillion within the next few years. The South Korean collaboration follows similar agreements Nvidia has recently secured in the United Arab Emirates, Saudi Arabia, Europe, and the United Kingdom.

During its second-quarter earnings report in August, Nvidia forecasted that sovereign AI projects could generate more than $20 billion in revenue for fiscal year 2026. The company’s strategy to secure such partnerships has become increasingly significant as trade tensions between the United States and China have severely restricted Nvidia’s operations in China, once one of its most lucrative markets.

Analyst Gil Luria from D.A. Davidson noted that the sovereign AI initiative is vital to Nvidia’s expansion, as the construction of global AI infrastructure underpins the company’s long-term growth. He explained that since China previously accounted for up to a quarter of Nvidia’s sales, the firm must now strengthen its presence in other international markets to maintain its momentum in AI data center technology.

As part of the South Korean expansion, Samsung, SK Group, and Hyundai each plan to establish AI factories that will operate with approximately 50,000 of Nvidia’s graphics processing units (GPUs). NAVER Cloud, often described as South Korea’s equivalent to Amazon Web Services, Google Cloud, or Microsoft Azure, will also enhance its AI infrastructure with 60,000 GPUs. The Ministry of Science and ICT intends to acquire more than 50,000 Nvidia chips in the coming years for use in the national AI computing center and through collaborations with companies such as NHN Cloud and NAVER Cloud. The government will further employ Nvidia’s software to support the development of its domestic AI models.

Raymond Teh, Nvidia’s Vice President of Sales and Marketing for the Asia-Pacific region, stated that the agreements will enable South Korean enterprises and institutions to establish a sovereign AI cloud within the country.

Hyundai plans to utilize Nvidia’s Blackwell GPUs to advance its autonomous driving and robotics programs, while SK Telecom, a subsidiary of SK Group, will integrate the new hardware to build industrial AI cloud infrastructure across Asia. Nvidia’s announcement in South Korea came shortly after U.S. President Donald Trump’s meeting with Chinese President Xi Jinping for high-level trade discussions.

Earlier in the week, Nvidia CEO Jensen Huang publicly praised the Trump administration and revealed a new U.S. sovereign AI project in collaboration with the Department of Energy, underscoring the company’s growing role in shaping national AI strategies worldwide.