NAVER, NVIDIA, and Brookfield to expand South Korea's AI infrastructure to 200 MW by 2028

The planned expansion will take place at NAVER's GAK Sejong data center in Sejong, South Korea, where the NVIDIA DSX platform will be deployed to provide a full-stack AI factory infrastructure. The initial buildout of 55 megawatts, announced previously, will be increased to 200 megawatts by 2028, with NAVER targeting a future scale of 1 gigawatt. This development aims to support AI innovators in Korea and the U.S. by offering large-scale AI compute resources for next-generation AI models and services.

NVIDIA has committed to investing $1 billion in NAVER, while Brookfield has signed a nonbinding term sheet to provide up to $9 billion in financing. NAVER will cover the remaining funding required for the project. NVIDIA's investment is contingent on customary closing conditions and NAVER securing at least $9 billion in committed financing separate from NVIDIA's contribution.

The expanded AI factory will incorporate advanced NVIDIA AI infrastructure, including the Vera Rubin and Blackwell platforms, designed to deliver efficient compute, software, and support for developing and deploying competitive AI applications at scale. NAVER, as an NVIDIA Cloud Partner, brings expertise in operating hyperscale infrastructure powered by NVIDIA's full-stack AI platform.

Brookfield, with approximately $100 billion in assets under management across AI infrastructure sectors such as data centers, compute, semiconductor manufacturing, and power generation, will leverage its investment capabilities to support this project. The NAVER investment represents a key milestone in Brookfield's strategic partnership with NVIDIA, combining capital and infrastructure expertise to advance large-scale AI factory deployments.

The NVIDIA DSX platform offers an integrated stack covering chips, systems, software, and facilities, optimized to reduce costs and accelerate AI production. Its DSX MaxLPS software enhances throughput per megawatt, while DSX OS manages lifecycle, health automation, resiliency, and multi-tenant operations across the expanded infrastructure.

In addition to infrastructure expansion, NAVER and NVIDIA continue to collaborate on open AI model development. NAVER is advancing its HyperCLOVA X models based on NVIDIA's Nemotron 3 Ultra open models, contributing proprietary data and training expertise. NAVER is also the first Korean member of the NVIDIA Nemotron Coalition, supporting open model development across various training stages.