Nvidia CEO Jensen Huang has stated that he will support the creation of an “Artificial Intelligence (AI) Valley” in Saemangeum, North Jeolla Province, modeled after Silicon Valley in the United States. This reflects his intention to position Korea, which can secure eco-friendly power and land, as an advance base for the era of physical AI.

Concluding his four-day visit to Korea on 8 June, Huang visited the headquarters of major domestic companies including SK, Hyundai Motor, LG, and Naver, seeking to draw key Korean players in semiconductors, robotics, and data center infrastructure into the Nvidia ecosystem.

● Jensen Huang: “Will participate in creating an AI Valley in Saemangeum”

On 8 June, at Hyundai Motor Group’s Yangjae headquarters in Seocho-gu, Seoul, Huang held a roughly one-hour discussion with Chung Euisun, chairman of Hyundai Motor Group. Speaking to reporters afterward, Huang said, “We talked about a new city that integrates AI into future manufacturing systems,” and added, “Just as there is Silicon Valley in the United States, an AI Valley is being created in Saemangeum, Korea.”

Huang also exchanged views with Chung on mobility and robotics. Huang said, “We are collaborating to accelerate robotic technologies that can be used in industrial sites,” while Chung stated, “It is an honor to now be working together on mobility and robotics with CEO Huang, with whom we formed a relationship 15 years ago.”

Earlier, at the LG Twin Towers in Yeouido, Huang had also discussed cooperation on physical AI with Koo Kwang-mo, chairman of LG Group. The two sides plan to expand strategic cooperation in areas such as building data, simulation, training, and behavior for next-generation robotics. Huang said, “From robotic systems to the A Factory, we are working with the LG Group as if we were one team across all of Nvidia’s business areas.” Koo added, “We had extensive discussions with Nvidia about our future direction.”

On 7 June, the previous day, Huang also met with Doosan Group Chairman Park Jeong-won to discuss AI cooperation plans, including the supply of power equipment by Doosan Enerbility for AI factories.

● Strengthening semiconductor and infrastructure collaboration in Korea

Huang has also been seeking allies in Korea in semiconductors and data centers, which form the foundation for building hardware. From 8:30 a.m. that day, he met with SK Group Chairman Chey Tae-won at the SK Seorin Building in Jongno-gu, Seoul, and agreed to upgrade the existing high-bandwidth memory (HBM) partnership between the two companies into one that encompasses everything from chip design and manufacturing to data center operations.

Huang expressed his confidence, saying, “SK hynix has been Nvidia’s largest memory partner and will continue to be so,” to which Chey responded, “We will elevate cooperation between the two companies to the group level so that we can build future AI factories together and share our research and development (R&D) roadmap.” Under this plan, SK hynix will jointly develop customized memory for Nvidia’s next-generation chipsets, and SK Telecom will work on building a gigawatt (GW)-scale AI cloud targeted to start operating in 2027.

In the afternoon, Huang visited Naver’s “1784” headquarters in Seongnam, Gyeonggi Province, where he reached a sudden agreement with Naver Executive Chairman of the Board Lee Hae-jin and CEO Choi Soo-yeon on a joint project to build a gigawatt (GW)-scale ultra-large AI factory. The development of a “Seoul World Model,” in which Nvidia’s next-generation world model “Cosmos” is combined with Naver’s Street View-based maps so that AI can precisely recognize real-world spaces, will also move into full swing.

Huang then held a working-level meeting at The Shilla Seoul in Jung-gu, Seoul, with Samsung Electronics Device Solutions (DS) Division head Jun Young-hyun (vice chairman) and other executives, and also met with Minister of Science and ICT Bae Jae-hoon. He concluded his four-day visit to Korea with a closed-door dinner at the same venue with the heads of domestic tech startups.

Experts analyzed that Nvidia is pursuing a “lock-in strategy” to fully incorporate Korea’s advanced manufacturing infrastructure into the Nvidia ecosystem. Just as the company previously distributed its proprietary software platform “CUDA” free of charge, effectively forcing AI developers to use Nvidia, it is now attempting to transplant the same successful approach to physical AI, a next-generation core industry. Lee Kyung-mook, professor at Seoul National University Business School, said, “It is positive that Nvidia, which leads the AI ecosystem, is seeking to secure dominance by cooperating with Korean companies,” but added, “However, we must guard against the risk that our companies could become dependent on Nvidia if their bargaining power weakens in the future.”

Popular News