In a sweeping partnership that signals South Korea’s emergence as a major AI hub, NVIDIA has committed to supplying 260,000 of its latest GPUs to the Korean government and the country’s leading tech conglomerates. The move comes as the chipmaker seeks to diversify beyond China amid escalating U.S.-China tech tensions, while positioning Seoul as a strategic anchor for its Asian operations.

NVIDIA CEO Jensen Huang announced the collaboration at the APEC CEO Summit in Gyeongju on October 31, detailing plans to deploy cutting-edge Blackwell GPUs and other chips across Korea’s public and private sectors. The initiative dovetails with South Korea’s ambition to rank among the world’s top three AI powers and represents one of the largest national AI infrastructure investments announced this year.

The GPU allocation breaks down as follows: the South Korean government will receive 50,000 units through the Ministry of Science and ICT, distributed across the National AI Computing Center and domestic cloud providers including NHN Cloud, Kakao Corp., and NAVER Cloud. On the corporate side, Samsung Electronics, SK Group, and Hyundai Motor Group will each secure 50,000 GPUs, while NAVER Cloud will obtain 60,000 units—the largest single allocation in the deal.

Samsung’s Semiconductor AI Factory

Samsung Electronics plans to harness its 50,000-GPU allocation to build what it calls a “semiconductor AI factory,” deploying NVIDIA’s Omniverse platform to create digital twins of its chip fabrication processes. The company is also tapping NVIDIA’s Cosmos and Isaac robotics platforms to accelerate development of next-generation home robots. Chairman Lee Jae-yong indicated Samsung aims to help set new industry standards through the partnership.

SK Group’s Industrial AI Cloud

SK Group will deploy more than 50,000 GPUs to power an AI factory focused on semiconductor R&D and cloud infrastructure. SK Telecom, the group’s telecom arm, will offer industrial cloud services using NVIDIA’s RTX PRO 6000 Blackwell Server Edition GPUs, giving domestic manufacturers access to infrastructure for digital twin and robotics projects.

Hyundai’s $3 Billion Physical AI Push

Hyundai Motor Group is committing roughly $3 billion alongside NVIDIA and the Korean government to build out a physical AI ecosystem. The automaker will use 50,000 Blackwell GPUs in an AI factory designed to train models for autonomous driving, smart manufacturing, and robotics at scale. Chairman Chung Euisun framed the partnership as essential to leading what he called the “AI-based mobility and smart factory era.”

NAVER’s Cloud Expansion

NAVER Cloud will take the lion’s share with 60,000-plus GPUs, positioning itself to operate Korea’s largest AI infrastructure while expanding services for enterprise and physical AI workloads. The company, which has commercialized its HyperCLOVA X model across search, translation, and automation services, expects the GPU boost to significantly improve training speeds and model performance.

Building Korea’s Sovereign AI

Beyond hardware, the Korean government is partnering with NVIDIA on a Sovereign AI Foundation Models initiative. LG AI Research, NAVER Cloud, NC AI, SK Telecom, and Upstage are collaborating to develop Korean-language large language models using NVIDIA’s NeMo and Nemotron frameworks, with plans to create AI agents featuring voice reasoning capabilities.

The partnership also extends to next-generation telecommunications. NVIDIA has signed agreements with Samsung Electronics, SK Telecom, KT, LG Uplus, Yonsei University, and the Electronics and Telecommunications Research Institute to jointly develop AI-RAN (AI radio access network) technology—a 6G advancement that promises to reduce power consumption and enhance computing efficiency by offloading GPU tasks to base stations.

In quantum computing, the Korea Institute of Science and Technology Information (KISTI) will work with NVIDIA to establish a Center of Excellence, leveraging the national supercomputer “Hangang” to build hybrid quantum computing capabilities. Using NVIDIA’s CUDA-Q platform and NVQLink architecture, researchers will focus on quantum error correction and hybrid application development.

Startup Ecosystem Support

NVIDIA is bolstering Korea’s startup scene through its Inception program, creating an alliance that connects emerging companies with GPU infrastructure from partners like SK Telecom and funding from venture firms including IMM Investment, Korea Investment Partners, and SBVA. Participating startups will gain access to NVIDIA’s software stack and technical expertise. The company is also joining the Ministry of SMEs and Startups’ N-Up AI incubation program.

Deputy Prime Minister and Science Minister Bae Kyung-hoon characterized the partnership as pivotal to South Korea’s transformation, noting that expanding AI infrastructure with NVIDIA will reinforce the country’s manufacturing strengths at a moment when AI has evolved from innovation to industrial foundation.

Huang, for his part, drew a parallel between Korea’s historic manufacturing prowess and its AI potential, suggesting the country could now “produce intelligence as a new export” to drive global transformation, just as its physical factories once amazed the world with ships, cars, and semiconductors.

Bloomberg noted the deal benefits both sides: NVIDIA gains a strategic foothold after being largely shut out of China, while Korean conglomerates secure stable access to chips that have become increasingly scarce amid a global AI infrastructure buildout expected to exceed $1 trillion.

The partnership will quadruple South Korea’s AI computing capacity from 65,000 to over 300,000 GPUs—a 4.5-fold expansion that represents far more than a hardware procurement deal. By encompassing infrastructure, R&D, talent development, and startup support, the NVIDIA-Korea collaboration marks a comprehensive bet on positioning the country as a frontline player in the global AI race.