One detail that surfaced later: SK Hynix has placed a 44.2 billion won capital expenditure to scale HBM4 production specifically for Nvidia's next architecture . No total dollar value for the partnership was disclosed by either company
.
The SK Group relationship expanded beyond memory into full-stack AI cloud services. SK Telecom announced plans to build a gigawatt-scale AI Cloud in South Korea using Nvidia's DSX (AI Factory) reference architecture, with the first AI factory planned to come online in 2027 .
This is not a pilot project. A gigawatt-scale deployment would make it one of Asia's largest planned AI infrastructure investments . The AI cloud is designed to support training, inference, and agentic workloads for AI startups, robotics, and industrial physical AI across South Korea, with eventual expansion into broader Asian markets
.
Crucially, the deal represents a deliberate broadening of the SK-Nvidia relationship. Where previous cooperation centered on SK Hynix supplying HBM, the new partnership extends into data center construction and operation, pulling SK Telecom's network and real estate assets directly into Nvidia's infrastructure roadmap .
Naver, South Korea's internet giant, signed a strategic alliance with Nvidia to build full-stack AI factories at gigawatt scale—explicitly framed around sovereign AI .
Naver described the partnership as spanning the "entire value chain," from exploring global demand to capital cooperation and joint responsibility for outcomes . The company is not just buying GPUs; it is building a complete AI factory platform using Nvidia DSX, integrating chips, servers, networking, and software
.
The buildout starts at 55 megawatts at the GAK Sejong data center and scales toward gigawatt capacity . Naver's goal, as stated by its executives, is to become a "global AI factory" serving enterprises and governments that want to run AI workloads on domestic infrastructure using domestic data
.
Huang himself named Naver Cloud a key partner in global AI infrastructure during GTC Taipei 2026 earlier that month, and Naver Chairman Lee Hae-jin attended the Seoul meetings in person—signaling the alliance's strategic weight .
Beyond the headline deals, Huang's meetings extended Nvidia's reach into Korean industrial conglomerates:
Multiple sources detail the partnership announcements extensively, but none report same-day U.S. stock market data for Nvidia shares or the Nasdaq 100 in response to the news. The announcements were made on Monday, June 8, 2026, in Seoul—which would have fallen during U.S. pre-market or early trading hours. Without specific trading data in the available evidence, any claim about how the market reacted would be speculation. A definitive answer would require checking U.S. market closing prices for that session.
Taken together, the deals mark a structural shift. Nvidia is no longer simply buying memory from South Korea; it is co-designing chips, co-building data centers, and co-developing sovereign AI factories with Korean partners. As Huang reportedly told the press in Seoul, "The build-out of AI infrastructure is only just beginning" . The agreements signed during this visit are designed to shape the rest of that buildout.
Comments
0 comments