NAVER, NVIDIA, and Brookfield Triple Korea's AI Factory Capacity, Signaling Sovereign AI Acceleration

NAVER, NVIDIA, and Brookfield have announced a substantial expansion of South Korea's sovereign AI factory infrastructure, a move set to significantly bolster the nation's AI capabilities. The initial NVIDIA DSX AI factory at NAVER's GAK Sejong data center is slated for a massive upgrade, increasing its capacity from 55 megawatts to 200 megawatts by 2028. NAVER has also articulated an ambitious long-term vision to further scale this deployment to a remarkable 1 gigawatt. This significant investment, unveiled during Korea President Jae Myung Lee's AI Summit visit, involves a $1 billion investment from NVIDIA and a commitment of up to $9 billion from Brookfield, with NAVER covering the remaining financial requirements. The expanded infrastructure will leverage cutting-edge NVIDIA AI platforms, including the forthcoming Vera Rubin and Blackwell architectures, ensuring state-of-the-art compute capabilities. This initiative is a profound indicator of the intensifying global competition to establish sovereign AI capabilities. For technical practitioners, it signals a crucial shift and diversification within the AI infrastructure landscape. Alongside the established global hyperscalers, national or regional "AI factories" are emerging as increasingly vital compute providers. This development holds particular importance for organizations grappling with stringent data governance, regulatory compliance, or national security mandates, as it presents viable alternatives to relying solely on foreign-owned cloud services. The sheer scale of the investment further solidifies the industry's consensus that advanced AI development necessitates immense, specialized hardware coupled with tightly integrated software stacks, moving beyond the capabilities of general-purpose compute. This directly impacts the availability, cost, and strategic control of high-performance AI compute resources within the APAC region and potentially influences global AI infrastructure strategies. This expansion aligns perfectly with the broader, well-established trend of nations and large enterprises striving for greater autonomy and control over their AI development and deployment. The concept of "sovereign AI" has rapidly gained momentum, fueled by escalating concerns regarding data residency, intellectual property protection, and geopolitical considerations. Governments and major corporations worldwide are increasingly channeling significant investments into dedicated AI infrastructure to cultivate domestic innovation ecosystems and mitigate dependence on a concentrated group of dominant global cloud providers. NVIDIA, through its comprehensive full-stack AI platform—encompassing DSX, Blackwell, and Vera Rubin—continues to play an indispensable role as a foundational enabler for these large-scale AI deployments, underscoring its strategic importance across the AI hardware and software spectrum. This move also reflects the ongoing capital expenditure arms race among leading technology companies and nations, all vying to secure the foundational compute necessary for the next wave of AI advancements. In practical terms, practitioners must recognize that the landscape of AI compute is undergoing significant diversification. While global hyperscalers will undoubtedly retain their importance, specialized AI factories like NAVER's GAK Sejong facility will offer distinct and compelling advantages for specific use cases. Developers and architects engaged in AI projects involving sensitive data or subject to stringent regulatory frameworks should proactively investigate the service offerings and operational models of such sovereign AI initiatives. This trend also implies a sustained and growing demand for expertise in managing, optimizing, and deploying workloads on NVIDIA's accelerated computing platforms. Organizations should closely monitor the pricing structures, access mechanisms, and support for open-source models provided by these emerging "AI factories," as they could present highly competitive alternatives for deploying large language models (LLMs) and other sophisticated AI applications. The long-term objective of achieving 1 gigawatt of capacity strongly indicates a sustained, strategic commitment to positioning Korea as a pivotal global AI hub, which is likely to attract further talent, investment, and innovation to the region.