NVIDIA and South Korea Partner on Sovereign AI at APEC CEO Summit

At the APEC CEO Summit, NVIDIA announced a strategic partnership with South Korea's public and private sectors to establish a robust sovereign AI infrastructure. This initiative involves deploying hundreds of thousands of NVIDIA GPUs to power AI factories and sovereign clouds, targeting critical sectors like automotive, manufacturing, and telecommunications.

"Korea’s leadership in technology positions it at the heart of the AI industrial revolution," stated Jensen Huang, founder and CEO of NVIDIA. He emphasized that accelerated computing is now as essential as power grids for national growth.

The South Korean government, represented by the Ministry of Science and ICT, plans to deploy 50,000 new NVIDIA GPUs to support research and business innovation. According to the original report titled "NVIDIA and South Korea align on sovereign AI at APEC CEO Summit", the first phase will feature 13,000 NVIDIA Blackwell GPUs integrated through providers such as NAVER Cloud and Kakao.

Strategic AI Factories and Industrial Impact

Major Korean corporations are scaling investments to align with national computing goals:

• Samsung: Partnering with NVIDIA to build an AI factory running over 50,000 GPUs. The facility will utilize NVIDIA Omniverse for digital twins and cuLitho to accelerate chip production.
• SK Group: Constructing a massive AI facility expected by 2027, offering GPU-as-a-service. SK Telecom is also developing the A.X. foundation model using NVIDIA AI Enterprise.
• Hyundai Motor Group: Investing $3 billion alongside NVIDIA to enhance physical AI capabilities, focusing on autonomous driving and factory automation via the DRIVE AGX Thor platform.

Sovereign AI Models and Research

To ensure digital sovereignty, NAVER Cloud is deploying 60,000 GPUs for industry-specific models. Simultaneously, the Ministry of Science and ICT is spearheading the Sovereign AI Foundation Models project using NVIDIA NeMo, collaborating with partners like LG AI Research and Upstage.

In the scientific domain, KISTI will integrate NVIDIA computing into the HANGANG supercomputer to bridge quantum processors with GPU clusters. Furthermore, the NVIDIA Inception program will foster a local startup alliance, providing emerging companies with the high-performance computing resources necessary to compete globally.