SK Group, NVIDIA sign $500B+ AI partnership spanning factories and memory

On July 24, 2026, SK Group and NVIDIA announced a $500-billion-plus comprehensive partnership spanning AI factory construction and AI memory supply, signaling a deepening strategic alignment between the South Korean conglomerate and the American AI computing leader. The two sides signed letters of intent to formalize the agreement, building on decades of prior technology partnerships.

The cornerstone of the deal is SK Telecom's commitment to build a 2-gigawatt AI Cloud in Korea using NVIDIA's DSX platform and NVIDIA Vera Rubin accelerated computing hardware, with memory supplied by SK hynix's HBM4 chips. The facility is planned to come online in 2027 and will be built on NVIDIA's DSX full-stack AI factory architecture, which integrates NVIDIA accelerated computing, systems, software, and partner technologies to deliver what the companies describe as the lowest token cost at maximum energy efficiency. In parallel, SK hynix is entering into a long-term AI memory partnership with NVIDIA to codevelop and optimize next-generation AI memory solutions, including HBM, to meet evolving demands ranging from large language model training to agentic AI and physical AI applications.

SK Group Chairman Chey Tae-won framed the partnership as a path for Korea to transcend its role as a leading AI adopter and become a global hub driving AI innovation, leveraging SK hynix's AI memory and SK Telecom's infrastructure capabilities. NVIDIA founder and CEO Jensen Huang echoed this positioning, stating that "South Korea has all the ingredients to become a global AI powerhouse — world-class networks and data centers, leadership in chip technology and vast industrial scale." The two companies aim to jointly accelerate large-scale AI infrastructure development and address increasing AI demand across the Asia-Pacific region, including applications in sovereign, physical, agentic, and enterprise AI services.

SK Group's partnership with NVIDIA represents a major commitment to making South Korea a hub for AI infrastructure rather than merely a consumer or adopter of AI technology. The collaboration links three complementary capabilities: SK Telecom's existing data-center and cloud-services expertise, SK hynix's foundry-level chip design and memory manufacturing (HBM4), and NVIDIA's computing platform and accelerator architecture. By anchoring a 2-gigawatt facility to launch in 2027, the companies are signaling a multi-year, capital-intensive bet on serving both local Korean demand and the wider Asia-Pacific region.

The partnership also addresses a structural constraint in AI infrastructure: SK hynix gains a committed, long-term customer for next-generation memory, while NVIDIA secures stable supply of the memory chips essential to its inference and training systems. SK Telecom, meanwhile, can offer cloud customers access to cutting-edge compute infrastructure at scale without bearing all capital risk alone. The combination of on-site memory manufacturing (SK hynix) and cloud deployment (SK Telecom) on a single platform (NVIDIA DSX) is intended to minimize token cost and energy consumption—metrics that matter to cloud operators facing rising power constraints.

No comments yet. Be the first to share your thoughts!