NVIDIA Partners with South Korea on Sovereign AI Initiatives at APEC CEO Summit

At the APEC CEO Summit, NVIDIA unveiled its ambitious initiative to collaborate with both public and private sectors in South Korea, aiming to establish a robust sovereign AI infrastructure. This groundbreaking plan encompasses the deployment of hundreds of thousands of NVIDIA GPUs within sovereign clouds and AI factories, targeting essential sectors such as automotive, manufacturing, and telecommunications.

A New Era of AI in Korea

Jensen Huang, the founder and CEO of NVIDIA, stated, "Korea’s leadership in technology and manufacturing positions it at the heart of the AI industrial revolution—where accelerated computing is as crucial as power grids and broadband." He noted that just as Korea has set benchmarks in manufacturing sophisticated ships, cars, chips, and electronics, it is now poised to export intelligence, marking a significant step towards global transformation.

Bae Kyung-hoon, Korea’s Deputy Prime Minister, echoed this sentiment, emphasizing that AI has transitioned from mere innovation to the bedrock of future industries, placing South Korea on the verge of a monumental transformation.

Government Initiatives and Support

The South Korean government is committed to launching up to 50,000 NVIDIA GPUs to enhance sovereign AI programs aimed at businesses and research groups. The initial phase will deploy 13,000 NVIDIA Blackwell and other GPUs through providers such as NAVER Cloud, NHN Cloud, and Kakao. This expansion includes plans to establish a National AI Computing Center, where startups, researchers, and organizations can harness this sovereign infrastructure for model training and application development.

NVIDIA’s collaborations extend to major companies like Samsung, SK Telecom, ETRI, KT, LGU+, and Yonsei University, focusing on AI-RAN and next-gen 6G networks. This work aims to offload some computing tasks from devices to network base stations, potentially reducing battery drain and cutting costs for sovereign AI services.

Major Companies Building Sovereign AI Factories

Strategic Investments in AI Infrastructure

Prominent corporations in Korea are pouring resources into advanced AI infrastructure, supporting national goals for sovereign computing capabilities. For instance, NVIDIA and Samsung are set to construct a state-of-the-art AI factory that seamlessly integrates chip manufacturing with accelerated computing. This facility will house more than 50,000 NVIDIA GPUs, enhancing data-driven production methods like predictive maintenance and optimizing processes across chip fabs.

Jensen Huang remarked, "We are at the dawn of the AI industrial revolution—an era set to redefine global design, construction, and manufacturing." Echoing this vision, Jay Y. Lee, executive chairman of Samsung Electronics, expressed enthusiasm for their ongoing partnership with NVIDIA, dating back to their collaboration on DRAM for NVIDIA’s graphics card in 1995.

Samsung plans to leverage NVIDIA CUDA-X libraries alongside software from Synopsys, Cadence, and Siemens to accelerate circuit design and manufacturing workflows. Additionally, by using NVIDIA Omniverse, Samsung will create digital twins of its factories and equipment, enhancing real-time simulation, testing, and logistics—paving the way for broader sovereign AI adoption.

Expanding AI Capacity: SK Group’s Vision

A Forward-Looking Approach

SK Group is ambitiously constructing an AI factory that will include over 50,000 NVIDIA GPUs, projected to be operational by late 2027. This facility is designed to offer GPU-as-a-service to support SK subsidiaries as well as external clients, harmonizing with South Korea’s sovereign AI strategy. Collaboratively with NVIDIA, SK Group is also developing next-generation high-bandwidth memory intended for GPUs.

Chairman Chey Tae-Won asserted, "SK Group is working with NVIDIA to position AI as a transformative force that transcends traditional limits in speed, scale, and precision across various industries."

SK Telecom aims to establish an industrial AI cloud featuring NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs. This platform will assist in semiconductor manufacturing, digital twins, and internal AI operations. Meanwhile, SK hynix is employing NVIDIA PhysicsNeMo tools for chip design simulations, targeting enhanced accuracy and efficiency.

To further support its workforce, SKT is developing an innovative foundation model named A.X., powered by NVIDIA NIM microservices and NVIDIA AI Enterprise, to assist numerous employees in chip development and operational functions.

Hyundai Motor Group’s AI Ambitions

Charting the Future of Transportation

NVIDIA and Hyundai Motor Group are expanding their partnership, focusing on autonomous driving, factory automation, and advanced robotics. Hyundai plans to construct an AI factory utilizing NVIDIA Blackwell GPUs to facilitate integrated training, simulation, and deployment processes.

Jensen Huang highlighted the transformative potential of AI across various industries, stating, "In transportation—from vehicle design to robotics—NVIDIA’s AI platforms are redefining how we move."

This partnership is expected to yield a joint investment of approximately $3 billion to bolster national physical AI capabilities. Their comprehensive plan encompasses establishing an NVIDIA AI Technology Center, Hyundai’s Physical AI Application Center, and new data centers aimed at nurturing the next generation of AI talent.

Navigating the Landscape of Sovereign AI Models

Innovative Initiatives from NAVER Cloud and Others

NAVER Cloud is set to introduce more than 60,000 GPUs for sovereign and physical AI applications, creating industry-specific models particularly in shipbuilding and public safety sectors. Concurrently, the Ministry of Science and ICT is spearheading a Sovereign AI Foundation Models project using NVIDIA NeMo and open Nemotron datasets, collaborating with entities such as LG AI Research and SK Telecom.

LG is actively engaging with NVIDIA on physical AI research while providing EXAONE models to support startups and researchers, particularly in healthcare applications.

Advancements in Quantum and Scientific Research

KISTI’s Innovative Vision

The Korea Institute of Science and Technology Information (KISTI) plans to leverage NVIDIA accelerated computing in its sixth national supercomputer, HANGANG. This initiative will facilitate NVQLink—an open architecture designed to connect quantum processors with GPU clusters, while also nurturing scientific foundation models and physics-informed AI tools with NVIDIA PhysicsNeMo.

In an exciting collaborative effort, NVIDIA and local partners are forming a startup alliance through the NVIDIA Inception program, opening the door for accelerated computing resources and funding support for aspiring entrepreneurs.

As South Korea propels itself into the AI-driven future, these collaborative efforts between government bodies and tech giants signal a profound transformation on the horizon. As you embark on your own journey of innovation, consider how you can leverage AI technologies to elevate your endeavors and make a lasting impact in your industry.