Pranav Hotkar
Seoul, South Korea - June 7, 2026 - NVIDIA and SK Hynix have announced a new multiyear technology partnership focused on developing next-generation AI memory and accelerating the buildout of global AI factory infrastructure, deepening collaboration between two of the most important companies in the rapidly expanding AI ecosystem.
The agreement centers on co-developing advanced memory technologies for NVIDIA’s future AI platforms, including Vera Rubin AI supercomputers, AI-focused CPUs, RTX Spark-powered AI PCs, and Jetson Thor robotics systems. The companies also plan to apply AI technologies to semiconductor design and manufacturing processes as they seek to scale infrastructure for increasingly complex AI workloads.
NVIDIA said the partnership aligns with the growing demand for AI factories, large-scale computing facilities optimized for AI training, inference, and autonomous AI systems. These environments require massive quantities of high-bandwidth memory (HBM), advanced networking, and high-density compute infrastructure.
“SK hynix and NVIDIA have been building toward this for years,” SK Group Chairman Chey Tae-won said in the announcement, describing the partnership as part of a broader effort to shape the future of AI infrastructure.
The companies said they will use NVIDIA CUDA-X libraries, PhysicsNeMo, and Omniverse technologies to accelerate semiconductor simulations, digital twin development, and factory optimization workflows. The initiative includes plans to create AI-powered digital twins for semiconductor manufacturing facilities to improve operational efficiency and automation.
The partnership builds on NVIDIA's and SK Hynix's longstanding collaboration around HBM memory used in AI accelerators. SK hynix currently holds a leading position in the global HBM market and remains NVIDIA’s largest memory supplier for AI infrastructure deployments.
Industry-wide, demand for AI memory has surged as hyperscale operators race to deploy larger GPU clusters and next-generation AI systems. NVIDIA CEO Jensen Huang recently warned that supply constraints across memory, silicon photonics, and advanced semiconductor infrastructure could continue for years because of accelerating AI demand.
The announcement also reflects South Korea’s broader ambitions to strengthen its role in the global AI infrastructure market. NVIDIA has recently expanded partnerships across the country involving AI factories, cloud infrastructure, robotics, and semiconductor manufacturing initiatives with companies including SK Group, SK Telecom, Samsung, LG, and Naver.
As AI infrastructure investment accelerates globally, advanced memory technologies are becoming increasingly critical to future data center performance. High-bandwidth memory plays a central role in supporting large-scale AI model training and inference by reducing data bottlenecks and improving GPU utilization within hyperscale AI environments.
The expanded NVIDIA–SK hynix partnership further underscores how semiconductor manufacturing, memory technology, and hyperscale AI infrastructure are becoming increasingly interconnected as the industry scales toward larger and more power-intensive AI deployments.
