NVIDIA and SK Group establish AI factory in South Korea

NVIDIA and SK Group announced a collaboration to develop an Artificial Intelligence (AI) factory in South Korea featuring over 50,000 NVIDIA GPUs. The facility aims to improve semiconductor research, digital twin technology, and AI agent development, facilitating advancements across numerous industries.

The partnership focuses on enhancing SK hynix High Bandwidth Memory (HBM) and next-generation memory solutions tailored for NVIDIA GPUs and telecommunications infrastructure. The AI factory, expected to be one of Korea’s largest, aims for completion in late 2027.

Supported by NVIDIA’s technologies, including CUDA-X and the PhysicsNeMo framework, SK hynix plans to boost chip design efficiency through AI-driven computer-aided design simulations. Additionally, the factory will enable the development of digital twins to optimize semiconductor production.

Jensen Huang, founder and CEO of NVIDIA, remarked, “SK Group is a vital memory technology partner, helping NVIDIA create the world’s most advanced GPU computing platforms that power global AI progress.” Chey Tae-Won, chairman of SK Group, expressed that the partnership will enable significant advancements in memory technology, robotics, and AI-driven solutions.

The AI factory infrastructure will also be accessible to developers participating in the Korea government’s Sovereign AI Foundation Models project, promoting further innovation in AI applications. Moreover, SK Telecom will establish an industrial AI cloud composed of NVIDIA RTX PRO 6000 Blackwell GPUs to support manufacturing leaders in advancing robotics and AI.

Overall, this initiative represents a strategic effort to enhance productivity and operational efficiency across Korea’s manufacturing sector, leveraging NVIDIA’s advanced computing technologies.