The South Korean tech giant is courting overseas clients as it builds an AI factory backed by Brookfield and Nvidia

Wed 11 Nov 2026 · 20:10 EST · 3 min read

In AI

Share

Naver, the company best known as South Korea’s dominant search engine, is now talking shop with a very different crowd.

The firm is engaging with more than 12 AI data center customers across North America and other regions.

What Naver is building

The customer outreach ties into a larger plan run through Naver Cloud, the company’s cloud computing arm. Naver calls the project an “AI factory.”

The idea is simple enough. Naver wants to rent out GPU computing power to outside companies and governments, with many clients sharing the same infrastructure.

The anchor site is GAK Sejong, Naver’s hyperscale data center in South Korea. The AI factory is designed to expand what that facility can already do.

The money behind it is substantial. Naver secured an infrastructure financing package of up to $9 billion from Brookfield Asset Management. After the Brookfield agreement, Nvidia finalized a $1 billion strategic equity investment in the project.

The capacity roadmap

The facility is set to start with 55 MW of capacity in the first half of 2027.

AI, tech, and the markets they move—in one daily briefing.

Daily. Free. Join 34,000+ readers across crypto, finance, and policy.

We respect your privacy. Unsubscribe anytime.

From there, the plan calls for scaling up to 200 MW. The longer-term target is 1 GW, reached as part of a phased overseas expansion strategy.

The hardware will come from Nvidia. The site is planned to use Nvidia’s next-generation DSX platforms, along with chips such as Blackwell and Vera Rubin.

Naver is not starting from zero, either. The company already runs GPU clusters, including 4,000 Nvidia B200 GPUs as of 2026. It also operates cloud regions in South Korea, Japan, Germany, Singapore and the US West Coast.

Why sovereign AI is the pitch

Naver’s stated target market has been enterprise and government clients in Asia, Europe and the Middle East. The focus is on sovereign and private AI workloads.

Naver Cloud is positioned as a regional competitor to US hyperscale providers, which collectively dominate global cloud computing. It is also responding to significant demand for AI data center capacity inside South Korea.

What this means

The most important word in Naver’s update is “engages.” Talking with more than a dozen potential customers is not the same as signing a dozen contracts.

That distinction matters because the AI factory is a capital-heavy bet. Financing of up to $9 billion and a $1 billion equity stake from Nvidia create expectations. The facility needs paying tenants to justify that spending, and conversations need to turn into commitments before the 55 MW phase comes online in 2027.

The next milestones to watch are straightforward. Watch for announcements converting these engagements into signed customers, any updates on the GAK Sejong construction schedule, and details on where the phased overseas expansion lands first.

Disclosure: This article was edited by Diego Almada Lopez. For more information on how we create and review content, see our Editorial Policy.