logo

On June 2nd, 2026, NAVER CLOUD CEO Kim Yu-won shared the direction of the Global AI Factory, a collaborative initiative with NVIDIA, at the NVIDIA Cloud Partner Summit (NCP Summit) held in Taiwan. The following summarizes key points from the speech.

​

​

NAVER CLOUD’s full-stack technological capability preparing for the AI factory era

​

The paradigm of the AI industry is shifting from model training toward an era of inference-focused AI factories, where the stable operation of large-scale infrastructure is critical. Just as power plants formed the foundation of the industrial era, AI factories will serve as the core bedrock of the new AI economy in this age of massive AI infrastructure.

​

NAVER CLOUD has long anticipated this shift, leveraging its comprehensive full-stack capability that span the entire AI lifecycle, from infrastructure to services. Rather than merely chasing industry trends, the company has focused strictly on building core technological competencies. From data centers and cloud services to foundation models and AI applications, NAVER CLOUD designs and operates the entire ecosystem with its proprietary technology, backed by a steady stream of investment driven by its commitment to Sovereign AI.

​

At the beginning of this journey was a close cooperation with NVIDIA. In 2020, NAVER CLOUD became the first company in the world to deploy NVIDIA DGX SuperPOD. Working through numerous trials and errors from the outset, the two companies accumulated deep, practical know-how in operating large-scale AI infrastructure. This experience successfully expanded into commercial success; today, NAVER CLOUD provides AI models and GPUSaas* to South Korea’s leading enterprises and institutions, including Hyundai Motors, Korea Hydro & Nuclear Power, and the Bank of Korea. Furthermore, the company is actively participating in building the “National AI Computing Center”, which is set to become the backbone of Korea’s AI industry.

​

*GPUaaS: Short for GPU-as-a-Service, refers to the delivery of GPU computing resources as a cloud service.

​

​

Expanding tech collaboration with NVIDIA across infrastructure, models, and physical AI

​

Building on years of collaboration and a shared foundation, NAVER and NVIDIA are further expanding their technological partnership across the entire AI ecosystem. NAVER CLOUD’s HyperCLOVA X is being enhanced by leveraging NVIDIA’s open large language model (LLM) technology, Nemotron 3 Ultra. Furthermore, both companies plan to jointly research core technologies and the optimization of hyperscale language models. This represents a true full-stack collaboration that spans the entire model development lifecycle, from data pre-processing to training and optimization.

​

Sovereign AI models begin with sovereign data. A prime example is the “Seoul World Model” which NAVER CLOUD developed by utilizing NVIDIA’s physical AI platform, Cosmos. Trained on South Korea’s national map data and 1.2 million panorama images captured from all across Seoul, the model successfully replicates the city’s actual road environments and the spatial structures. Moving forward, NAVER CLOUD aims to continuously expand its collaboration with NVIDIA on Cosmos-based physical AI models.

​

​

NAVER CLOUD’s leap to become Asia’s AI infrastructure hub

​

The partnership with NVIDIA extends far beyond a conventional GPU provider-customer relationship; it is a strategic decision to co-develop AI technologies and expand the global AI ecosystem. To this end, NAVER CLOUD aims to position itself as a key provider driving the explosive AI demand in the Asian market and as an unparalleled AI infrastructure hub. Furthermore, by steadily scaling up its infrastructure, the company plans to evolve into a gigawatt (GW)-scale global AI factory operator over the mid-to-long term.

​

NAVER CLOUD’s mission is clear: to empower users all around the world translate AI infrastructure into tangible, real-world value. Together with NVIDIA, NAVER CLOUD is committed to laying the groundwork for the upcoming Age of Intelligence ●

​

​

► Read more on Chairman Lee Hae-jin and Jensen Huang’s Announcement on ‘GW-Scale AI Factory’ Global Partnership at 1784 Visit (Link)