Nvidia has announced two significant investments in South Korea's artificial intelligence infrastructure, committing $1 billion to internet giant Naver and forging a $500 billion commercial partnership with conglomerate SK Group. The moves underscore the chipmaker's aggressive expansion into sovereign AI capabilities and its reliance on South Korean memory chip suppliers.
Investment in Naver's AI Data Center
Nvidia will invest $1 billion in Naver, the operator of South Korea's dominant search engine and a leading cloud platform, to help finance an AI data center currently under construction. Originally planned for 55 megawatts of power capacity, the facility will now be expanded to 200 megawatts, more than tripling its size, thanks to the Nvidia funding. US private equity firm Brookfield Asset Management has signed a nonbinding term sheet to provide up to $9 billion for the project, which will run entirely on Nvidia's AI computing hardware.
The data center is part of Naver's broader strategy to host its own large language models and AI services, reducing reliance on foreign cloud providers. Naver already operates one of the world's largest Korean-language AI models, and the expanded facility will support both internal workloads and external customers requiring high-performance AI compute. The investment follows a pattern where Nvidia injects capital into data center operators that subsequently purchase its GPU systems, a structure critics describe as circular financing.
Massive Partnership with SK Group
Separately, Nvidia announced a long-term commercial partnership with SK Group valued at more than $500 billion. This figure includes Nvidia's purchases of high-bandwidth memory (HBM) chips from SK Hynix, the world's largest supplier of that critical component, as well as acquisitions by SK Group of Nvidia's supercomputers. SK Telecom, the group's telecom arm, will build over 2 gigawatts of AI data centers on the Korean Peninsula, with the first facility scheduled to open next year.
The partnership also includes a design collaboration: Nvidia will help SK Hynix engineer future generations of HBM memory. This is a strategic move to secure supply of the advanced memory chips that are essential for training and running large AI models. HBM is in short supply globally due to the explosive buildout of AI data centers, and SK Hynix currently holds a commanding market share. The deal gives SK Hynix a structural advantage over its domestic rival Samsung Electronics, the only other company capable of producing HBM at scale.
Geopolitical and Economic Context
Nvidia CEO Jensen Huang described South Korea as being in a "golden age," citing the country's booming semiconductor and industrial sectors as reasons for the heavy investment. The announcements coincided with a visit to Silicon Valley by South Korean President Lee Jae Myung, and additional tech deals between Korean and American companies are reportedly in the works. The investments reflect a broader trend where countries seek sovereign AI infrastructure—independent data centers and models controlled within national borders.
South Korea is particularly well positioned for this shift. It possesses world-class chip fabrication capabilities, abundant engineering talent, and a strong government push to lead in AI. Samsung and SK Hynix are the only two firms globally capable of producing the advanced HBM memory that Nvidia's processors require. By locking SK Hynix into a design partnership, Nvidia ensures a dedicated supply line while simultaneously supporting the country's economic ambitions.
Nvidia's Global Investment Spree
The Korea deals are the latest in a year-long investment spree that has seen Nvidia commit more than $40 billion to AI equity positions, including $30 billion in OpenAI. Earlier this week, the chipmaker signed a separate agreement with Amkor Technology to bolster chip packaging capacity. It recently completed a trip to Japan that resulted in partnerships with Toyota and several robotics firms. Critics have noted that many of these investments involve Nvidia putting money into companies that then become customers for its hardware—a circular model that effectively builds a captive market for its GPUs.
Despite the controversy, the strategy has proven effective in expanding Nvidia's ecosystem. By financing the construction of data centers and the development of AI platforms, Nvidia ensures that demand for its chips remains robust even as geopolitical tensions threaten supply chains. The company is betting that the global buildout of AI infrastructure will require sustained investment, and South Korea, with its deep semiconductor roots, is a key battleground.
Implications for the AI Memory Market
The design partnership between Nvidia and SK Hynix could reshape the HBM market. Currently, HBM3E and future HBM4 chips are critical for Nvidia's upcoming Blackwell and Rubin GPU architectures. By collaborating on chip design, SK Hynix can tailor its memory to Nvidia's specific performance and power requirements, potentially widening its lead over Samsung. This comes at a time when Samsung is struggling to qualify its HBM3E products for Nvidia's latest processors, though recent reports suggest it has made progress.
Analysts believe that the partnership will accelerate the development of next-generation HBM, with larger capacities and higher bandwidth, enabling more efficient AI model training. The Korean government has welcomed the deal, seeing it as a vote of confidence in its semiconductor strategy. President Lee has prioritized AI and chip manufacturing as pillars of economic growth, and the Nvidia investments align with that vision.
Broader Industry Trends
Nvidia's push into South Korea is part of a wider phenomenon known as sovereign AI. Huang has argued that every country should create its own AI capability to ensure data sovereignty and economic competitiveness. This philosophy has driven Nvidia to invest in data center projects across Europe, the Middle East, and Southeast Asia. South Korea, with its unique linguistic and cultural context, is an ideal candidate for a Korean-language AI ecosystem, which Naver aims to lead.
The circular financing model has drawn scrutiny from regulators and industry observers. Some worry that it concentrates market power in Nvidia's hands, potentially stifling competition among cloud providers and hardware manufacturers. Others see it as a pragmatic approach to jumpstart the AI industry in regions that lack the capital to build infrastructure independently. For now, Nvidia's investments are fueling rapid expansion, and South Korea stands to gain both economically and technologically.
As the global AI race intensifies, the cooperation between Nvidia and South Korean companies will likely deepen. The $1 billion Naver investment and $500 billion SK partnership are just the beginning. With the first SK Telecom data center set to open next year and Naver's expanded facility on the horizon, South Korea is positioning itself as a hub for AI innovation, anchored by the world's most valuable chipmaker.