NAVER targets 55MW NVIDIA DSX AI factory at GAK Sejong

From data center expansion to AI factory strategy

NAVER is extending its AI infrastructure plans from cloud capacity into sovereign AI factory territory. NVIDIA Newsroom posted on June 8 that NAVER is building a full-stack NVIDIA AI factory, adding that GAK Sejong is planned for expansion to 55 megawatts with a longer-term path toward gigawatt scale. FxTwitter data shows 42,343 views, 368 likes, and three attached images.

The linked NVIDIA Newsroom release says the expansion will use NVIDIA DSX, a platform described as a full AI factory blueprint spanning chips, systems, software, facilities, and partner technologies. NAVER plans to use the stack for next-generation HyperCLOVA X models, its Seoul World Model work, agentic AI services, and AI cloud offerings for Korean and global customers.

The technical stake is cost and control. Frontier and regional AI systems now depend on post-training, high-volume inference, reliability automation, and token cost per megawatt, not only on raw GPU count. A 55MW starting point at GAK Sejong gives NAVER a concrete domestic base for large-scale AI services while positioning the company to sell sovereign infrastructure to customers that care about data locality and regulatory control.

NVIDIA Newsroom usually publishes official company updates, partnership notes, and infrastructure strategy items, so this post is best read as a partner roadmap rather than a casual social clip. The next things to watch are the deployment schedule at GAK Sejong, whether HyperCLOVA X improvements appear in measurable enterprise products, and whether NAVER’s AI cloud can convert sovereign AI demand in Korea, Europe, and the Middle East into durable usage. The source post is on NVIDIA Newsroom’s X account.