NVIDIA has emerged as the unifying force behind South Korea’s national push to industrialize artificial intelligence. In partnership with Samsung Electronics, SK Group, Hyundai Motor Group, and NAVER, NVIDIA is powering a wave of AI factories—each designed to train and deploy large-scale models for manufacturing, robotics, mobility, and telecommunications. Together, these projects represent tens of billions in private investment and the creation of one of the world’s most concentrated clusters of GPU infrastructure.
At the APEC Summit, NVIDIA and Samsung announced plans for a semiconductor AI factory featuring more than 50,000 NVIDIA GPUs. The facility will integrate CUDA-X, cuLitho, and Omniverse platforms to accelerate chip design and fabrication through AI-driven simulation and digital-twin environments. Samsung will apply NVIDIA’s software stack—including Nemotron, Cosmos, and Isaac GR00T—to robotics and factory automation, while deploying RTX PRO 6000 Blackwell GPUs for real-time logistics and autonomous fab operations. The collaboration builds on a 25-year alliance that began with DRAM for NVIDIA’s first graphics card and now extends deep into AI-optimized semiconductor manufacturing.
SK Group is also investing heavily in NVIDIA-powered infrastructure, building an AI factory with more than 50,000 GPUs to support SK hynix, SK Telecom, and external developers. SK hynix is deploying PhysicsNeMo and CUDA-X to speed chip simulation and digital-twin modeling of fabs, while SK Telecom is launching an industrial AI cloud using RTX PRO 6000 Blackwell GPUs to host robotics and manufacturing workloads. The SK–NVIDIA partnership also expands into advanced memory co-development for next-generation GPU platforms and future 6G telecommunications infrastructure.
Hyundai Motor Group is collaborating with NVIDIA and the Korean government to build an AI factory powered by 50,000 NVIDIA Blackwell GPUs. The system will train, validate, and deploy AI models for autonomous vehicles, smart factories, and robotics. Hyundai will use NVIDIA DRIVE AGX Thor for in-vehicle intelligence and Omniverse for simulating factory operations—forming a unified ecosystem where cars, robots, and production systems learn from shared data. NAVER, meanwhile, plans to expand its NVIDIA AI infrastructure with an additional 60,000 GPUs to develop sovereign and physical AI workloads tailored to Korean industries such as shipbuilding, logistics, and public services.
The Korean government, through the Ministry of Science and ICT (MSIT), is coordinating a national effort around NVIDIA’s technology stack. The Sovereign AI Foundation Models project—led by MSIT with partners LG AI Research, NAVER Cloud, SK Telecom, Upstage, and NVIDIA—aims to produce Korean-language foundation models using NeMo software and Nemotron datasets. KISTI, Korea’s national research institute, is using NVIDIA CUDA-Q and the PhysicsNeMo framework on its HANGANG-6 supercomputer to explore hybrid quantum computing and physics-informed AI.
NVIDIA is also partnering with Korean telecoms—Samsung, SK Telecom, KT, LGU+, and ETRI—to develop AI-RAN and 6G architectures that use GPU acceleration at base stations to reduce device power consumption and latency. To sustain the ecosystem, NVIDIA is expanding its Inception program in Korea with new startup alliances, venture capital partnerships, and a Center of Excellence powered by RTX PRO 6000 Blackwell GPUs. The NVIDIA Deep Learning Institute will support workforce training in AI, robotics, and quantum-accelerated computing.
• NVIDIA anchors Korea’s AI transformation with over 220,000 GPUs across Samsung, SK, Hyundai, and NAVER projects
• Samsung building semiconductor AI factory using CUDA-X, cuLitho, and Omniverse for digital-twin fabs
• SK Group constructing GPU-as-a-service AI factory for SK hynix and SK Telecom, advancing HBM and memory R&D
• Hyundai Motor Group deploying 50,000 Blackwell GPUs for mobility, robotics, and manufacturing
• NAVER adding 60,000 GPUs for sovereign AI models and public-service applications
• Government-backed Sovereign AI initiative developing Korean foundation models with NVIDIA NeMo and Nemotron
• KISTI using CUDA-Q and PhysicsNeMo for hybrid quantum research; 6G AI-RAN collaboration underway with major telcos
“We are at the dawn of the AI industrial revolution — a new era that will redefine how the world designs, builds, and manufactures,” said Jensen Huang, founder and CEO of NVIDIA. “Korea’s industrial leaders are building their AI foundations with NVIDIA to power the next generation of intelligent factories, mobility, and national innovation.”
🌐 Analysis: NVIDIA’s influence in Korea extends beyond chip supply—it is embedding its full technology stack into the country’s manufacturing, telecom, and research fabric. Korea’s coordinated AI-factory strategy places NVIDIA at the center of a national transformation, establishing a sovereign infrastructure model that could shape global industrial policy. With more than 220,000 GPUs planned across four corporate ecosystems, Korea is becoming a proving ground for the world’s first AI-industrial economy.
