Nvidia is turning a major South Korea visit into a broader industrial AI push, tying its chips and software platforms to manufacturing, mobility, telecom networks, cloud services, and robotics.
During this week's APEC Summit 2025, Nvidia CEO Jensen Huang visited South Korea for the first time in 15 years as the company and the South Korean government announced an expanded partnership. The plan is built around AI infrastructure and physical AI, with Korean companies set to receive the largest share of a major GPU allocation.
A national AI buildout with industry at the center
South Korea will secure over 260,000 of Nvidia's latest GPUs to support rising AI demand, according to the South Korean government. The distribution shows how the country is trying to connect national AI capacity with the companies that make semiconductors, vehicles, cloud services, and industrial systems.
Around 50,000 GPUs will support public initiatives. Those include the development of domestic AI foundation models and a national AI data center.
The remaining over 200,000 GPUs will go to companies such as Samsung, SK, Hyundai Motor Group, and Naver. The goal is to support AI-based manufacturing innovation and industry-specific AI model development, rather than treating AI infrastructure as a standalone computing project.
The timing also matters. The announcement came just days after the U.S. signed technology deals with Japan and South Korea covering emerging technologies including AI, semiconductors, quantum computing, biotech, and 6G.
Samsung brings AI into production and networks
Samsung is one of the clearest examples of how Nvidia's South Korea partnerships connect AI to industrial operations. Samsung announced plans to build an AI mega-factory with Nvidia, using more than 50,000 Nvidia GPUs and the Omniverse platform.
The facility is intended to bring AI into every stage of manufacturing for semiconductors, mobile devices, and robotics. Its intelligent network will be designed to analyze, predict, and optimize production in real time.
Samsung and Nvidia have been partners for over 25 years. Their work is now expanding to HBM4, a next-generation memory designed to power future AI applications.
The partnership also extends into telecom infrastructure. Nvidia will work with Samsung, SK Telecom, KT, LG Uplus, and ETRI (Electronics and Telecommunications Research Institute) to co-develop AI-RAN, according to the Korean Ministry of Science and ICT.
AI-RAN combines mobile base stations with AI to boost performance and cut battery use. Under the new agreement, Nvidia and South Korea's industry and research institutions will jointly develop next-generation AI-RAN and a global testbed, the Korean government stated.
Hyundai links physical AI to mobility and factories
Hyundai and Nvidia are focusing on physical AI, autonomous mobility, smart factories, and robotics. Their partnership also includes collaboration on high-performance GPU supply and investment.
According to Nvidia, the companies will use 50,000 NVIDIA Blackwell GPUs for integrated AI model training, validation, and deployment. They also plan to establish AI research centers in South Korea to strengthen the country's physical AI industry.
“AI is revolutionizing every facet of every industry, and in transportation alone — from vehicle design and manufacturing to robotics and autonomous driving — Nvidia’s AI and computing platforms are transforming how the world moves,” said Huang. “Together with Hyundai Motor Group — Korea’s industrial powerhouse and one of the world’s top mobility solutions providers — we’re building intelligent cars and factories that will shape the future of the multitrillion-dollar mobility industry.”
The Hyundai work shows why physical AI is a central phrase in the announcement. The focus is not only on software models, but on systems that operate inside factories, vehicles, robotics, and industrial environments.
SK and Naver push AI into cloud and real-world systems
SK Group, the parent of SK Hynix, is partnering with Nvidia to build Asia's first enterprise-led manufacturing AI cloud. The project will use Nvidia's simulation and digital twin platforms.
Access is planned for the government, public institutions, and domestic startups. That makes the SK effort more than an internal corporate deployment; it is positioned as shared infrastructure for industrial AI work.
Naver Cloud, the cloud arm of Korean search engine Naver, is working with Nvidia on a next-generation “Physical AI” platform. The platform is intended to connect the physical and digital worlds.
Naver Cloud plans to deploy AI infrastructure across key industries including semiconductors, shipbuilding, energy, and biotechnology. According to Naver, the aim is to accelerate the adoption of AI solutions optimized for real-world industrial environments.
“Just as the automotive industry is transitioning to SDVs, the era of ‘Physical AI,’ where AI operates directly within real industrial sites and systems, is unfolding,” Hae-jin Lee, founder of Naver, said in Naver's statement.
What the partnerships signal
Nvidia's collaborations with Hyundai Motor, Samsung, SK, and Naver point to a wider shift in how AI infrastructure is being deployed. The focus is moving from general computing capacity toward systems tied directly to hardware, factories, vehicles, telecom networks, and cloud platforms.
The announcement also sits inside a broader run of Nvidia partnerships. Earlier this week, the company announced new work with companies including Eli Lilly, Palantir, Hyundai, Samsung, Uber, and Joby Aviation, along with the U.S. Department of Energy.
Huang also sought to downplay concerns about an AI bubble. The news sent Nvidia's stock soaring, as Nvidia became the first publicly traded company to surpass a $5 trillion market capitalization.
For South Korea, the core message is practical: AI infrastructure is being tied to national models, data centers, manufacturing, mobility, telecom networks, cloud systems, and industry-specific applications. For Nvidia, the South Korea push shows how its role in AI is expanding beyond chips into the industrial systems those chips help run.