Seoul, Oct 31 (NationPress) Nvidia is set to roll out as many as 260,000 graphics processing units (GPUs) in South Korea, collaborating with the government and key corporations to construct extensive artificial intelligence (AI) factories across the nation, according to an announcement made by the US tech giant on Friday.

This initiative will see the involvement of the South Korean government along with major industry players such as Samsung Electronics, SK Group, Hyundai Motor Group, and Naver Cloud Corp.. Each of these entities will establish dedicated AI computing centers that will utilize Nvidia's cutting-edge Blackwell GPUs, as reported by Yonhap news agency.

Nvidia revealed in a preliminary media briefing on Thursday that the Seoul government plans to allocate up to 50,000 Nvidia GPUs to facilitate the development of a national “sovereign AI” platform aimed at training large language models (LLMs) optimized for the Korean language and various industrial applications.

GPUs, or graphics processing units, are high-performance semiconductor chips adept at executing multiple calculations simultaneously, making them essential for training and executing AI models.

Each of the major players—Samsung, SK, and Hyundai—will implement up to 50,000 GPUs, while Naver Cloud is expected to deploy approximately 60,000 units, culminating in a total of 260,000 GPUs.

Nvidia's announcement indicates that the infrastructure based on Blackwell technology will increase South Korea's total installed AI GPU capacity from roughly 65,000 to over 300,000 units, establishing the country as one of the largest centers for AI computing outside the United States.

This news comes as Jensen Huang, Nvidia's CEO, visits South Korea to participate in the Asia-Pacific Economic Cooperation (APEC) CEO Summit in Gyeongju, the venue for this year’s gathering of regional leaders.

Huang arrived in South Korea on Thursday, where he met with Euisun Chung, Executive Chair of Hyundai Motor Group, and Lee Jae-yong, Chairman of Samsung Electronics, in Seoul.

The Nvidia chief stated that his visit was in response to an invitation from President Lee Jae Myung. The two leaders convened in Gyeongju to discuss matters related to AI and economic collaboration.

Details regarding the financial aspects of these arrangements have not been disclosed.

Samsung plans to utilize the new AI factory to enhance semiconductor production through digital twin simulations and advanced robotics, leveraging Nvidia's NeMo, CUDA-X, and Omniverse platforms.

SK Group is expected to build a similar facility aimed at supporting semiconductor research and cloud-based AI applications, with SK Telecom Co. providing sovereign industrial cloud services utilizing Nvidia's RTX Pro 6000 Blackwell Server Edition GPUs.

Hyundai Motor intends to deploy the 50,000 Blackwell GPUs for training large AI models focused on autonomous driving, robotics, and smart factory operations, alongside plans to co-invest approximately US$3 billion with the government to create a “physical AI” ecosystem.