Technology · AI
Nvidia Takes 4.5% Stake in Naver With $1 Billion Investment
The chip giant and South Korea's internet leader are building a $10 billion AI infrastructure hub at Naver's Sejong data center to anchor the nation's sovereign AI ambitions.

KEY TAKEAWAYS
- ·Nvidia is investing $1 billion in Naver for approximately 4.5% ownership, making it the third-largest shareholder if the deal closes.
- ·The two firms are partnering with Brookfield on a $10 billion AI infrastructure project centered on Naver's Sejong data center in South Korea.
- ·The initiative aims to establish South Korea's sovereign AI factory, reducing dependence on foreign hyperscale cloud providers for domestic AI workloads.
A Strategic Stake in South Korea's Cloud Leader
Nvidia is putting $1 billion into Naver, the South Korean internet and cloud services company, through a third-party share issuance that will give the chip maker approximately 4.5% ownership. If the transaction closes and Naver's treasury shares are retired, Nvidia would become the company's third-largest shareholder, cementing a deeper relationship between the two firms at a moment when Asia's governments are racing to build domestic AI capacity.
The investment underscores Nvidia's willingness to take equity positions in strategic partners rather than simply selling chips. Naver operates one of South Korea's largest web portals and has been expanding its cloud infrastructure and AI research capabilities, making it a natural anchor tenant for Nvidia's data center ambitions in the region.
Building a $10 Billion AI Factory
Beyond the equity stake, Nvidia and Naver are joining forces with Brookfield, a global investment management firm, on an AI infrastructure plan valued at up to $10 billion. The centerpiece of that effort is Naver's GAK Sejong data center, which the partners intend to transform into the nucleus of South Korea's sovereign AI factory initiative.
Sovereign AI has become a rallying cry across Asia, as governments seek to ensure domestic control over the compute resources, training data, and models that underpin language processing, industrial automation, and national security applications. South Korea has been particularly vocal about building indigenous AI capabilities, wary of depending entirely on hyperscale clouds headquartered in the United States.
The Sejong facility sits in a planned city south of Seoul that already houses government ministries and research institutions. Locating the AI factory there sends a clear signal that the project is meant to serve public as well as private interests, offering compute capacity for universities, startups, and agencies working on Korean-language models and sector-specific applications.
Why Brookfield Matters
Brookfield's participation brings infrastructure finance expertise and patient capital. Data centers require enormous upfront investment in real estate, power systems, and cooling before they generate revenue, and sovereign AI projects often carry policy goals that extend timelines beyond what venture investors tolerate. Brookfield has been deploying capital into digital infrastructure globally, betting that AI workloads will drive decades of demand for purpose-built facilities.
The three-way partnership distributes risk and aligns incentives. Nvidia supplies chips and networking gear, Naver contributes its existing data center footprint and cloud software stack, and Brookfield underwrites the construction and expansion phases. For Nvidia, the arrangement locks in a large, multi-year customer while advancing a national AI agenda that could open doors with other Korean enterprises and government buyers.
Regional Context and Competition
South Korea is not alone in pursuing sovereign AI. Japan has announced partnerships with domestic cloud providers and chip firms, Singapore is expanding its national supercomputing resources, and India is planning a large-scale AI infrastructure program. Each country wants to avoid the bottleneck that occurs when all advanced AI workloads must route through a handful of foreign hyperscalers.
Nvidia's investment in Naver positions the company at the center of South Korea's effort, but it also raises questions about how sovereign these projects truly are when the underlying silicon and software frameworks remain dominated by a single American vendor. Seoul will need to balance its desire for self-reliance with the practical reality that Nvidia's GPUs and CUDA ecosystem are the de facto standard for training large models.
The deal also puts pressure on rivals. AMD has been courting Asian cloud providers with more open software and competitive pricing, while local semiconductor firms are exploring ways to design AI accelerators tailored to specific workloads. Nvidia's equity stake in Naver makes it harder for competitors to displace its hardware at the Sejong facility, at least in the near term.
What Happens Next
The transaction still requires regulatory approval and the consent of Naver's existing shareholders. Assuming it clears those hurdles, construction and equipment procurement for the expanded Sejong data center will likely begin in the second half of the year. Naver has not disclosed a timeline for bringing the first phase of the AI factory online, but industry observers expect initial capacity to be available within eighteen to twenty-four months.
Longer term, the success of the Sejong project will hinge on whether it attracts a critical mass of users beyond Naver's own services. If Korean enterprises, research labs, and government agencies find the facility offers competitive performance and data residency guarantees, it could become a template for other sovereign AI initiatives across Asia. If adoption lags, the project risks becoming an expensive showcase with limited commercial traction.
For now, Nvidia's billion-dollar bet signals confidence that South Korea's AI ambitions are more than rhetoric and that Naver has the operational chops to execute at scale.
RELATED STORIES
Spot something wrong? Email editor@briefasia.com. We log every correction publicly.



