Nvidia just deepened its ties with South Korea’s tech and industrial giants — signing a string of partnerships that stretch from memory chips and AI data centers to robotics, mobility and industrial AI. The deals were announced during CEO Jensen Huang’s visit to Seoul, where he met leaders from SK Hynix, Naver, SK Telecom, Doosan, LG, Hyundai and Samsung. Financial terms weren’t disclosed, but the scope and strategic weight are clear.
What was announced
- SK Hynix — A multi-year technology partnership focused on advanced memory products for global AI data centers. Nvidia called SK Hynix its largest memory partner and said it already buys “billions of dollars” of product annually, with purchases set to rise. Huang warned SK Hynix’s planned doubling of memory wafer capacity by 2030 won’t be enough to meet AI demand. The partnership runs for more than two years with extension options. Analysts say the deal underscores growing demand for customer-specific, high-bandwidth memory for AI infrastructure.
- SK Telecom — Will build a gigawatt-scale AI cloud in South Korea using Nvidia technology, with the first AI data center expected online in 2027.
- Naver and Doosan — Both will use Nvidia tech for AI data-center projects. Doosan, which builds robots and supplies materials used in Nvidia’s Blackwell chips, also plans energy solutions to support Nvidia platforms.
- LG Group — Collaboration on electronics, mechanical systems and humanoid-robot AI, and joint discussions on next-gen data center architecture (cooling, power delivery, design).
- Hyundai Motor Group — Nvidia will deepen work on autonomous mobility, robotics and AI-driven manufacturing. Huang praised Hyundai’s planned AI data center in Saemangeum, calling the area an “AI Valley.”
- Samsung — Nvidia and Samsung executives discussed next-gen foundry chips, autonomous-driving chips, HBM5 memory and shipments of Groq’s AI LP30 chips (Samsung said those are slated for H2 this year).
Context and market reaction
- South Korea remains a global manufacturing hub for chips, electronics, autos and ships; SK Hynix and Samsung are the world’s top memory-chip makers.
- The country’s Kospi index had rallied over the prior six months on AI optimism but plunged after U.S. jobs data revived rate-hike fears. The Kospi closed down 8.3% on Monday; Samsung fell 10.2% and SK Hynix 7.7%. Huang dismissed concerns, saying investors should be “very excited” and that lower prices present buying opportunities.
- Separately, South Korea’s tech ministry plans a 2026 state AI project worth 2.08 trillion won that will secure 9,704 GPUs, including 2,016 Nvidia Vera Rubin GPUs.
Why this matters for crypto and blockchain audiences
- GPUs and AI data-center capacity are central to the broader compute economy. While data-center GPUs (like the ones Nvidia sells for AI workloads) are different from consumer GPUs used by many crypto miners, rising public and private investment in AI infrastructure can tighten global GPU supply and influence pricing and availability across markets.
- Expanded AI clouds and specialized memory for high-bandwidth workloads could accelerate on-chain and off-chain AI integrations, oracle/ML services, and other crypto projects that rely on large-scale model training or inference.
- Government procurement of thousands of GPUs signals national-level prioritization of AI compute capacity — a trend that could reshape where and how compute-heavy decentralized and centralized applications deploy their workloads.
Bottom line
Nvidia’s South Korea push ties its chip and AI stack to some of the country’s largest manufacturers and cloud players, reinforcing its supply chain and data-center ambitions. For crypto stakeholders, the bigger takeaway is that global AI compute is scaling fast — and that shift will have ripple effects on GPU markets, cloud compute availability, and the emergence of AI-enabled blockchain services.
Read more AI-generated news on: undefined/news
