NVIDIA CEO Jensen Huang Outlines AI Vision at APEC 2025 Deepening Partnership with South Korea

At APEC 2025, NVIDIA CEO Jensen Huang details his AI vision, including AI factories and key partnerships with Samsung and SK Hynix in South Korea.
NVIDIA CEO Jensen Huang Outlines AI Vision at APEC 2025 Deepening Partnership with South Korea

NVIDIA CEO Jensen Huang's AI Vision at APEC 2025 Deepens Partnership with Korea

With the APEC 2025 Summit in Gyeongju, South Korea, Jensen Huang, CEO of NVIDIA, addressed the press to outline the company's artificial intelligence picture and announce important collaborations within the country. With this being Huang's first visit in 15 years, it indicates how South Korea is increasingly seen as an important player in the world technology sector.

AI as a New Computing Platform

Huang stated that AI is not merely like a tool, such as a chatbot, but is a whole new computing platform. The AI platform allows computers to learn, train themselves, and perform very demanding tasks, requiring tremendous computing power supported by a solid software stack and infrastructure.

He went on to argue that computing has not fundamentally changed much for the past 60 years. Now accelerated computing, he said, is changing every layer of the industry. Now AI has ceased to be a mere tool. AI is "the work itself", able to invent industries, able to add to GDP growth.

South Korea's Part in Creating AI Factories

One dominant theme was "AI factories", which are actually infrastructure on a grand scale to produce the intelligence to be distributed through all industries. Huang described these factories as a newly defined utility such as the electric grid or the internet.

He singled out South Korea as a "special country", one that is uniquely placed for this era through its combination of technology, software, and manufacturing capabilities. To this end, NVIDIA had plans to partner with the South Korean government alongside major corporations like Samsung, SK Hynix, Hyundai, and Naver to supply over 260,000 GPUs toward the formation of an AI infrastructure in that country.

The Essentials from the Q&A Session

On Physical AI and Robotics in Korea

In its answer to the question on Physical AI meaning AI that understands the physical laws of nature like gravity and momentum Huang cited it to be a domain where Korea can excel. To name a few industries, it includes autonomous driving, robotics, shipbuilding, and semiconductor manufacturing. He pointed out Korea's very unique ability not just to manufacture robots but also to use them skillfully thus creating a niche for growth.

On Attracting a Global AI Ecosystem

Huang is optimistic that with the building of a local AI factory hosting 260,000 GPUs, South Korea will become the center of attraction for regional AI development. This facility will not only fuel local researchers and startups but also draw in outside companies to cement Korea's spot as one of the largest AI centers in the world.

On Samsung's Role and HBM4 Standards

Huang knew that Samsung produces NVIDIA's Orin processor for robotics and autonomous driving. He also expressed "great confidence" that all HBM suppliers, including those in Korea, will measure up to NVIDIA's standards for the next-gen Vera Rubin platform which should utilize HBM4 and HBM4E.

On Competition from Huawei

When competition from Chinese firms was mentioned, Huang said that Huawei is undoubtedly a stiff competitor and is well aware of 5G and smartphone technology. He said, "it would be foolish to underestimate competitors like Huawei" and asserted that NVIDIA is taking the competition quite seriously. "NVIDIA is running faster as China continues to catch up with us in no time", he added.

On Uncontrollable AI

When asked about fears from experts about an AI that can't be controlled, Huang replied simply "It's science fiction".

NVIDIA's Core Value Extreme Co-design

Huang's last words were that the foremost value attested to NVIDIA is "Extreme Co-design". Essentially, what this means is that it does not design a single chip or algorithm; it actually designs the entire data center stack together from CPU, GPU, system architecture, software, and AI models. This holistic optimization, he argued, is a capability that no other company can pull off with NVIDIA's scale, thus enabling rapid innovation and the delivery of new capabilities every year.

About the author

mgtid
Owner of Technetbook | 10+ Years of Expertise in Technology | Seasoned Writer, Designer, and Programmer | Specialist in In-Depth Tech Reviews and Industry Insights | Passionate about Driving Innovation and Educating the Tech Community Technetbook

Join the conversation