International Business: Nvidia secures AI infrastructure deals with South Korean giants
Nvidia said it signed multi-year AI infrastructure deals in South Korea with SK Hynix, SK Telecom, Naver and Doosan, without disclosing deal values. Nvidia and SK Group said SK Hynix will develop advanced memory for global AI data centers; SK Telecom plans a gigawatt-scale AI cloud using Nvidia tech, with a first data center in 2027. Nvidia CEO Jensen Huang also met LG and Hyundai. South Korea’s ministry plans 9,704 GPUs for a 2.08 trillion won state AI project in 2026, including 2,016 Vera Rubi

Deal flow reinforces Nvidia’s AI platform demand (GPUs + data-center systems) and supports memory supply continuity for its AI roadmap.
Nvidia announced multi-year AI infrastructure deals in South Korea with SK Hynix, SK Telecom, Naver, Doosan, LG and Hyundai, expanding supply and customer adoption.
Bullish bias for NVDA as the market prices in sustained AI infrastructure buildouts and tighter supply-chain execution.
Background
The article describes Nvidia’s CEO trip to South Korea and a set of announced AI infrastructure partnerships aimed at securing advanced memory and expanding data-center deployments.
Why it matters
The newest concrete items are the multi-year SK Hynix technology partnership and the stated plans for SK Telecom/Naver/Doosan to build AI data centers using Nvidia technology, plus the South Korea ministry’s GPU procurement plan including Nvidia’s Vera Rubin GPUs.
Market relevance
For traders, the key takeaway is incremental evidence of Nvidia’s expanding AI infrastructure ecosystem in South Korea (memory supply + data-center buildouts) alongside a government GPU procurement plan that includes Nvidia’s Vera Rubin GPUs.
Market effects
Reinforces the shift of memory from commodity to customer-specific AI supply arrangements and highlights data-center buildout demand for GPUs and supporting infrastructure.
South Korea’s AI supply chain (memory + telecom + cloud/data-center buildouts) is positioned as a key execution hub for Nvidia’s AI expansion.
Signals continued global AI infrastructure capex and supply-chain coordination between GPU platforms and advanced memory for data centers.
Alternative perspectives
Without disclosed deal values or incremental capacity commitments beyond stated partnerships, the market may already be pricing in Nvidia’s AI demand and treat this as incremental rather than earnings-moving.
Memory supply constraints are central; if SK Hynix’s capacity ramp (wafer doubling by 2030) remains slower than AI demand, Nvidia’s partners could still face bottlenecks despite partnership language.
Key entities
- companyNvidia
Announced multiple South Korea AI infrastructure deals and partnerships to secure memory supply and expand data-center buildouts.
- companySK Hynix
Signed a multi-year technology partnership with Nvidia to develop advanced memory for global AI data centers.
- companySK Telecom
Plans a gigawatt-scale AI cloud in South Korea using Nvidia technology, with first data center online in 2027.
- companyNaver
Will use Nvidia technology to help build AI data centers (per Nvidia statement).
- companyDoosan Group
Expected its energy solution to be used in Nvidia data-center platforms and to use Nvidia’s physical AI technology.



