NAVER Cloud stresses flexibility and deep expertise in tailoring AI workloads to customer needs, emphasizing sovereignty, security, and governance with dedicated and private cloud solutions integrated with AI technologies for global customers.

NAVER Cloud has previously enhanced its Seoul World Model with NVIDIA’s Cosmos platform, training on 1.2 million panoramic Seoul images to improve realism of road environments and spatial structures.

A strategic partnership between NAVER Cloud and NVIDIA aims to advance HyperCLOVA X using NVIDIA’s open-source Nemotron 3 Ultra LLM, with a focus on expediting large-scale AI infrastructure and global AI factory capabilities.

Dr. Kim Yuwon, CEO of NAVER Cloud, said the company provides end-to-end AI capabilities to build, train, and run AI, backed by a multi-region global network spanning Japan, Singapore, Germany, and the US, and a leading share in Korea’s public sector.

NAVER GIO Lee Hae-jin and NVIDIA CEO Jensen Huang are slated to meet in Korea soon to finalize execution plans for the global AI factory business.

NAVER Cloud plans to leverage its cloud capabilities to support Sovereign AI tailored for governments and enterprises, addressing rising AI demand in Asia with a flexible, client-specific AI environment.

Dr. Kim envisions a highly flexible, powerful AI cloud geared to customers prioritizing sovereignty and regulatory compliance, signaling a broad global Sovereign AI strategy across multiple markets.

The partners will conduct joint research on super large language models and support the development of sovereign AI models for the respective countries involved.

This collaboration emphasizes both global collaboration and national AI sovereignty through joint LLM research and country-specific sovereign AI model development.

The collaboration aims to shift AI development from model-centric approaches to inference-driven AI factories capable of handling massive-scale inference workloads and large-scale data processing.

The scope includes physical AI applications using NVIDIA’s Cosmos platform to enhance NAVER Cloud’s Seoul World Model, which replicates Seoul’s road environments with training on 1.2 million panoramic images.

This collaboration illustrates ongoing joint use of Cosmos data to boost NAVER Cloud’s Seoul World Model.