NVIDIA Corporation (NVDA.US) deepens its presence in the South Korean AI market: assists Samsung, SK, Hyundai in establishing AI factories. Samsung may supply HBM4 chips for it.

During the visit of Huang Renxun, CEO of NVIDIA Corporation (NVDA.US) in South Korea, on October 31st, NVIDIA Corporation announced partnerships with Samsung Electronics, SK Group, and Hyundai Motor Group to promote the construction of AI factories. According to the agreements reached, NVIDIA Corporation will supply over 260,000 accelerator chips to assist South Korea in launching AI projects. Samsung Electronics NVIDIA Corporation will collaborate with Samsung to build a new generation AI factory, marking a new era of integration of intelligent computing and chip manufacturing. Samsung's semiconductor AI factory will be equipped with over 50,000 NVIDIA Corporation graphics processing units (GPUs). Samsung has stated that they are in "close discussions" with NVIDIA Corporation regarding the supply of the next-generation high-bandwidth memory chip HBM4. The Korean company planning to launch the new chips next year did not disclose the specific shipping time for the HBM4 chips. Samsung's HBM chips are in competition with fellow Korean company SK Hynix, which is currently the main supplier of NVIDIA Corporation's HBM chips. American Micron Technology, Inc. (MU.US) is also a strong competitor in this field. SK Hynix plans to start shipping the latest HBM4 chips in the fourth quarter of this year and expand sales next year. In a joint statement about the AI factory, NVIDIA Corporation noted, "From our first collaboration using Samsung DRAM for NVIDIA Corporation's first graphics card NV1, to the launch of the first commercial HBM chip in the industry, and now the crucial supply partnership for HBM3E and HBM4, the solid 25-year alliance between the two companies has laid the foundation for today's AI revolution." Furthermore, Samsung is utilizing NVIDIA Corporation's CUDA GPU acceleration infrastructure to optimize chip manufacturing for the optical proximity correction platform, achieving a 20-fold performance improvement in computational lithography and technology computer-aided design simulation fields. Samsung is also using NVIDIA Corporation's Omniverse platform to build digital twin systems for global semiconductor factories, shortening the design-to-operations cycle, enabling AI-driven predictive maintenance, operational optimization, and real-time decision-making. Digital twins are a technology that involves virtual simulation of physical objects, processes, or systems using real-time sensor data. Samsung emphasizes that they are experiencing a significant leap in efficiency in the simulation, verification, and manufacturing analysis fields using NVIDIA Corporation's GPU, CUDA-X libraries, as well as solutions from Synopsys, Inc. (SNPS.US), Cadence, and Siemens. SK Group NVIDIA Corporation stated that they are partnering with SK Group to jointly build an AI factory to promote semiconductor research and development, cloud infrastructure development, support digital twin and AI intelligent body development. The AI factory developed by SK Group will be equipped with over 50,000 NVIDIA Corporation GPUs, with the first phase of the project planned to be completed by the end of 2027. The new factory will provide support to SK Hynix, SK Telecom, and other group subsidiaries, as well as external institutions through a GPU-as-a-service model. The two parties are also collaborating to develop SK Hynix HBM and next-generation advanced storage solutions for NVIDIA Corporation's GPUs, focusing on the semiconductor manufacturing and communication infrastructure sectors. In addition, SK Telecom, a cloud service provider for NVIDIA Corporation, plans to deploy an industrial AI cloud platform in Asia with NVIDIA Corporation's RTX PRO 6000 Blackwell server GPU, accelerating digital twin and Siasun Robot&Automation technology innovation for startups and government agencies. The initial deployment will include over 2,000 NVIDIA Corporation RTX PRO 6000 Blackwell GPUs running NVIDIA Corporation Omniverse workloads, supporting SK Magnachip Semiconductor Corp. manufacturing, wafer digital twin, and internal AI intelligent body applications. SK Telecom is developing a basic model called A.X., which is built on NVIDIA Corporation's NIM microservices and AI Enterprise software platform, empowering SK Hynix's AI intelligent body systems to help over 40,000 employees and production workers collaborate and accelerate problem-solving, improving chip research and development efficiency. Hyundai Motor Group NVIDIA Corporation and Hyundai Motor Group will collaborate on the new Blackwell architecture AI factory to promote autonomous driving vehicles, intelligent factories, and Siasun Robot & Automation technology innovation. The two parties will deploy 50,000 NVIDIA Corporation Blackwell GPUs to jointly develop AI capabilities for mobile travel solutions, build the next-generation intelligent factories, and advance equipment-side semiconductor technology breakthroughs. Integrating AI model training, verification, and deployment through GPU clusters is a core goal for both parties. To respond to South Korea's government initiative to build a national physical AI cluster, Hyundai Motor and NVIDIA Corporation will work together with government agencies to accelerate ecosystem development, with a total investment of approximately $3 billion to drive South Korea's physical AI development. The South Korean government has announced a broad collaboration with NVIDIA Corporation, procuring over 260,000 state-of-the-art AI GPUs for deployment in public and private sectors, including projects with Samsung, SK Group, and Hyundai Motor. NVIDIA Corporation and Hyundai Motor will jointly establish an AI Technology Center, Hyundai Motor Group's Physical AI Application Center, and a regional data center. According to the cooperation plan announced earlier this year, Hyundai Motor will use NVIDIA Corporation's Omniverse Enterprise platform to create factory digital twins. Hyundai Motor is currently exploring the use of NVIDIA Corporation's RTX PRO 6000 Blackwell server GPUs to develop factory digital twins and Siasun Robot & Automation systems using the Omniverse and Cosmos platforms. In addition, Hyundai Motor is developing advanced AI models based on NVIDIA Corporation's Nemotron open-source AI inference model and NeMo software platform, aiming to achieve cloud-based online upgrade capabilities for vehicle functions and features.