0
Naver and Nvidia push for a joint project to build a gigawatt-scale, ultra-large global AI factory. / Photo via Naver

Naver is teaming up with Nvidia to pursue a joint business project aimed at building a gigawatt-scale (GW), ultra-large global AI factory.

According to Naver on the 8th, Naver Founder and Chief Investment Officer Lee Hae-jin and Nvidia CEO Jensen Huang plan to meet at Naver's 1784 headquarters in Seongnam, Gyeonggi Province, on this day to discuss a specific roadmap for the partnership currently being driven by both companies, alongside detailed cooperation plans for jointly entering the global market. This follows an agreement on a strategic direction where both companies will secure leadership in the AI infrastructure ecosystem, extending beyond the Asia-Pacific region into the European and Middle Eastern markets.

This alliance is an integrated partnership that goes beyond simple technological cooperation, cutting across all stages of the value chain from global demand discovery to capital cooperation. Naver will participate as a core global partner, sharing joint responsibility for the project's performance and risks. To preempt exploding global AI demand, both companies plan to swiftly push forward with the construction of an ultra-large AI factory that will set the standard for global AI infrastructure, beginning with the groundbreaking of a 55-megawatt (MW) facility in 2027.

First, Naver will utilize its core hub, the hyperscale data center 'GAK Sejong,' as its outpost. Signaling a start with the operation of 55MW in the first half of 2027, the company plans to absorb global demand by expanding its infrastructure scale overseas to 100MW within the same year and up to 200MW by 2028. One gigawatt is a scale equivalent to approximately four times the maximum capacity of 'GAK Sejong,' which is Naver's largest hyperscale data center in South Korea. It is also an overwhelming size capable of simultaneously accommodating hundreds of thousands of Nvidia's latest graphics processing units (GPUs).

Both companies also plan to advance their technological ties across all fronts. Naver's accumulated capabilities in building and operating large-scale proprietary GPU clusters, combined with its hyperscale data center know-how, will be fully integrated with Nvidia's next-generation high-performance infrastructure platform, 'DSX.' Through this, they plan to maximize data center operational efficiency and significantly boost commercial viability. Furthermore, next-generation technological cooperation in the field of spatial intelligence will begin in earnest, such as building a 'Seoul World Model' utilizing Naver's proprietary spatial modeling and Street View data for Nvidia's world foundation model, 'Cosmos.'

Naver recently became the first South Korean company to join the 'Nvidia Nemotron Alliance,' which includes 12 global top-tier AI enterprises such as Cursor, Mistral AI, and Perplexity. Based on this, Naver is accelerating efforts to enhance the performance of 'HyperCLOVA X' and secure its global versatility by combining Nemotron's joint technological development achievements with Naver's proprietary data and accumulated training know-how.

Founder Lee Hae-jin stated, "It is highly encouraging that through this alliance, we can present a concrete alternative for various regions and nations around the world to build their independent Sovereign AI capabilities," adding, "This cooperation holds great significance in that it has provided an opportunity for the competitiveness of Naver's technological infrastructure to take a leap forward into the global market."

                                                                                                             Kim Young-jin