Samsung Electronics, SK Group and Naver chairmen are set to meet Nvidia CEO Jensen Huang near Silicon Valley on July 24, with OpenAI and Anthropic CEOs expected to join, in a gathering that could reshape AI supply chain dynamics across memory chips, foundry services and data center infrastructure.
The leaders of South Korea's three largest technology conglomerates will gather with Nvidia Corp. Chief Executive Jensen Huang near Silicon Valley on Friday, in what would be the broadest assembly of AI supply chain executives since the industry's infrastructure buildout accelerated last year.
Samsung Electronics Chairman Lee Jae-yong, SK Group Chairman Chey Tae-won and Naver Corp. Chairman Lee Hae-jin are expected to attend a roundtable with Huang, according to people familiar with the plans. OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei may also participate, though none of the companies has confirmed the guest list. South Korean President Lee Jae-myung is also visiting Silicon Valley this week as part of a five-nation trip.
"This meeting signals that Korean companies have moved beyond being memory suppliers to become core enablers of the AI infrastructure ecosystem," said Rachel Kim, a semiconductor supply chain analyst. "The scope of discussion has expanded from HBM procurement to foundry services, AI factory construction and model-level collaboration."
The gathering comes just six weeks after Huang spent four days in Seoul meeting with the same Korean conglomerates, during which Nvidia disclosed expanded agreements with each. This week's talks are expected to focus on translating those commitments into concrete procurement plans, manufacturing contracts and infrastructure investment timelines.
Memory and Foundry Expand Beyond Traditional Roles
Samsung Electronics and SK hynix Inc., the world's two largest memory chipmakers, already supply Nvidia with high-bandwidth memory (HBM — a type of DRAM that stacks chips vertically to achieve much higher data transfer speeds than conventional memory). Both companies' HBM4 products are slated for inclusion in Nvidia's next-generation AI accelerator, Vera Rubin, which entered full production in late May and is expected for market launch in the second half of this year.
The Korean memory makers also supply OpenAI's proprietary AI chips and signed an agreement last year to provide high-performance memory for Stargate, a multibillion-dollar AI infrastructure project.
Beyond memory, Samsung's foundry business — contract manufacturing of chips designed by other companies — is gaining traction within Nvidia's ecosystem. Samsung already produces Nvidia's Grok 3 LPU (language processing unit, a chip optimized for running AI inference rather than training), which pairs with Vera Rubin systems. Reports citing The Information said Anthropic is in early talks to use Samsung's 2-nanometer process (a manufacturing technology that packs more transistors per square millimeter, improving performance per watt) for its own AI chips. Neither company has confirmed the discussions.
SK Group has the most to protect from the talks. Its memory co-development deal with Nvidia widened in June to span four platforms covering AI infrastructure, personal AI and physical AI. SK Telecom Co. plans a gigawatt-scale AI cloud service in South Korea using Nvidia's DGX platform. Chey said in June that the group aims to finish its first domestic AI factory in 2027 and build in Japan later in the decade.
AI Factories and the Infrastructure Race
Naver, South Korea's largest internet company, is earlier in the same business. The company plans to construct AI factories that pair graphics processing units with data center space, power and cooling, then sell the resulting computing capacity as a service. Under the plan Nvidia disclosed in June, Naver will expand its Gak Sejong data center to 55 megawatts and eventually toward gigawatt scale. Naver has also joined Nvidia's Nemotron Coalition for open model development as its first Korean member.
Capacity that large needs anchor tenants. OpenAI and Anthropic are the sort of customers who could fill it, which is why their possible attendance matters beyond the guest list. If confirmed, the meeting would bring together the leaders of the world's two largest memory chipmakers, the dominant GPU platform provider and three of the most prominent generative AI developers — a full-stack alignment that industry observers say could signal a new phase of strategic coordination across the AI value chain.
The summit also arrives at a moment of market anxiety. Memory and chip stocks have fallen by double digits from their recent highs, and investors have questioned whether AI infrastructure spending can sustain its current pace. HBM supply constraints are projected to persist into 2027, according to industry estimates. A concrete investment roadmap emerging from this week's talks could help ease those concerns.
For Samsung, successful foundry deals with Nvidia and Anthropic would mean benefiting from both advanced logic manufacturing and HBM demand growth simultaneously — a dual revenue stream that could narrow the gap with Taiwan Semiconductor Manufacturing Co., which dominates the contract chipmaking market. For SK Group and Naver, securing long-term GPU supply commitments and AI factory customers would validate their multibillion-dollar infrastructure bets. Nvidia, meanwhile, gains deeper integration with Korean partners that are increasingly essential to its AI platform's memory supply, chip manufacturing and data center capacity.
This article is for informational purposes only and does not constitute investment advice.
