NAVER, Brookfield, NVIDIA expand Korea AI factory to 200 megawatts

On July 25, 2026, NAVER, Brookfield, and NVIDIA announced a major expansion of Korea's sovereign AI factory infrastructure. The centerpiece is a scaling up of the initial NVIDIA DSX AI factory deployment from 55 megawatts to 200 megawatts at NAVER's GAK Sejong hyperscale data center in Sejong, South Korea. The announcement came during Korea President Jae Myung Lee's AI Summit visit to San Francisco, highlighting the strategic importance of the initiative to South Korea's AI ambitions.

The $10 billion（約1.6兆円） project will be financed through a three-way partnership: Brookfield will provide up to $9 billion（約1.4兆円） as the exclusive capital partner, NVIDIA will invest $1 billion（約1600億円）, and NAVER will fund the remaining amount. This funding structure pairs Brookfield's global infrastructure capital with NVIDIA's DSX computing platform and NAVER's operational expertise in hyperscale data centers. According to Haejin Lee, Founder and Chairman of NAVER, the deal marks a transition from vision to "robust execution phase" for NAVER's AI Factory business, combining Brookfield's capital with NVIDIA's computing platform.

The expanded 200-megawatt facility will feature NVIDIA Vera Rubin and Blackwell platforms and is designed to serve both Korean and U.S.-based AI innovators. It will function as a dedicated resource pool for emerging AI companies, providing compute, software, and support to develop and deploy competitive AI models and applications at scale. The infrastructure also supports NAVER's broader ambition, announced in June, to extend the GAK Sejong data center with a long-term path to gigawatt-scale sovereign AI infrastructure serving Korea's enterprises, industries, government organizations, and global AI cloud customers.

Beyond infrastructure, the partnership deepens NAVER and NVIDIA's collaboration on open model development. NAVER is advancing its HyperCLOVA X models to be based on NVIDIA Nemotron 3 Ultra open models, leveraging its proprietary data and training expertise. NAVER has also become the first Korean company to join the NVIDIA Nemotron Coalition, contributing to open model development across pretraining, post-training, and reinforcement learning. As an NVIDIA Cloud Partner, NAVER brings deep expertise in operating hyperscale infrastructure powered by NVIDIA's full-stack AI platform.

The announcement represents a major step in South Korea's push to build sovereign AI infrastructure independent of foreign cloud giants. By partnering with Brookfield—a global infrastructure investor—and NVIDIA, NAVER is securing both the capital and the cutting-edge computing platform needed to scale from the initial 55-megawatt deployment to 200 megawatts. The deal was announced during Korea President Jae Myung Lee's AI Summit visit to San Francisco, underscoring the strategic importance Seoul places on AI competitiveness.

The three-way funding split—with Brookfield as the dominant capital partner, NVIDIA as both investor and technology provider, and NAVER covering the remaining share—aligns incentives across infrastructure development, semiconductor supply, and operator expertise. By positioning the facility as an open resource pool for emerging AI companies in both Korea and the U.S., NAVER is creating a regional alternative to hyperscalers that focus primarily on serving their own cloud services. The explicit mention of a longer-term path to gigawatt-scale infrastructure signals this is a foundation for much larger buildout.

NAVER's collaboration with NVIDIA on open model development, including its role as the first Korean company in the Nemotron Coalition, shows the partnership extends beyond raw compute: NAVER is helping shape the open-source AI model ecosystem while advancing its own HyperCLOVA X models. This positions South Korea as a center for both AI infrastructure and model innovation.

No comments yet. Be the first to share your thoughts!