Naver is expanding the AI infrastructure of its Sejong data center, 'Gak Sejong,' based on the NVIDIA DSX platform and is pursuing the construction of a gigawatt-class AI factory in the long term. The initial construction scale is 55 MW, and Naver plans to utilize this as an AI cloud base for the enterprise, industrial, and public sectors.

NVIDIA and Naver announced on the 8th that Naver is expanding its Sovereign AI infrastructure by adopting the NVIDIA DSX platform. This infrastructure is scheduled to be used for the advancement of HyperClova X, AI agent services, and the development of physical AI models.

The new infrastructure will be built around Naver's hyperscale data center, 'Gak Sejong'. Naver plans to meet the demand for large-scale AI model training and inference by combining the data center operational capabilities of Gak Sejong with NVIDIA's AI infrastructure platform.

Naver is enhancing HyperClova X using the NVIDIA Nemotron model and is also pursuing the development of the 'Seoul World Model' based on the NVIDIA Cosmos World Foundation model. In the second half of this year, an AI agent platform utilizing NVIDIA Nemoclo Blueprint is also scheduled to be introduced in Korea.

This collaboration comes amidst growing demand for sovereign AI in the global market. Naver is expanding its AI infrastructure business in Europe and the Middle East, while strengthening its AI cloud business to respond to local regulations and demands for data sovereignty.

NVIDIA DSX is an integrated platform for building AI factories that provides chips, systems, software, and data center infrastructure design together. Based on this, Naver plans to accelerate the speed of AI infrastructure deployment and operational efficiency, and expand its services to domestic and international AI cloud customers.