Naver announced on Monday its collaboration with Nvidia to establish a gigawatt-scale artificial intelligence factory, beginning at its hyperscale data center located in Sejong. This bold initiative was outlined during a meeting between Naver's founder and board Chairman Lee Hae-jin, CEO Choi Soo-yeon, and Nvidia's CEO Jensen Huang at Naver's headquarters in Seongnam, south of Seoul.
During the visit, Huang expressed his enthusiasm for Naver, stating, "I love Naver," and participated in a Naver Webtoon event where he humorously noted, "Don't worry! I have GPUs!"
The partnership, which is set to encompass more than just technological collaboration, aims to cultivate demand, facilitate investment, and develop the necessary infrastructure within the AI value chain. As a central ally, Naver will share both the financial outcomes and risks of the project.
The ambitious initiative will kick off at the Gak Sejong data center, where Naver intends to deploy 55 megawatts of AI infrastructure by the first half of 2027. The capacity will grow to 100 megawatts later that same year and ultimately reach 200 megawatts by 2028, paving the way for the planned gigawatt-scale AI operations. The facility is expected to meet AI demand not only within South Korea but also across Asia, the Middle East, and Europe.
Collaboration will focus on merging Naver’s expertise in data center operations and GPU cluster management with Nvidia's DSX platform, which combines chips, servers, software, and data center technology tailored for AI factories. This platform aims to reduce the costs associated with AI model training and inference while accelerating infrastructure deployment.
Naver is leveraging this technology to enhance its AI infrastructure services for various sectors, including businesses, governments, and industrial clients. The partnership is also geared toward broadening technical collaboration. Naver has been advancing its HyperCLOVA X AI model by integrating Nvidia's open large language model, Nemotron, and is developing a "Seoul world model" by blending Nvidia's Cosmos foundation model with its own street-view and spatial modeling technologies.
Future expansions in this partnership may lead to innovations in physical AI, robotics, and digital twins. This meeting represents a development in Naver and Nvidia’s ongoing partnership, which began last year when Lee and Huang convened at the Asia-Pacific Economic Cooperation summit in Gyeongju to explore collaboration on physical AI platforms.
Naver's 1784 facility is viewed as a showcase of advancing robotics, digital twins, and cloud innovations, and Huang's visit was an opportunity to assess stronger collaboration avenues. Naver aims to transition from its current focus on internet services to becoming a prominent global AI infrastructure provider. At Nvidia's GTC Taipei 2026, Huang highlighted Naver Cloud as a vital participant in the global AI landscape, while Naver signaled plans to enhance its sovereign AI and data center operations in partnership with Nvidia.
Following the announcement, Naver's stock experienced a significant uptick, closing at 279,000 won (approximately $181), marking a 9.20% increase from the last trading session, as market analysts linked this rise to heightened investor optimism surrounding the AI factory initiative and Naver’s broader AI infrastructure expansion efforts.
