NAVER Secures About $1.4 Billion in AI Factory Compute Contracts

By Elena Vasquez

NAVER (035420) has secured about 2 trillion won ($1.4 billion) in computing-service supply contracts for the AI factory it is building with Brookfield and Nvidia. Revenue is scheduled to begin when the service launches in the first half of next year.

The company disclosed the contracts during a third-quarter earnings conference call on the 12th. CEO Choi Soo-yeon said they cover an AI inference company and exceed the volume NAVER initially targeted.

“We have signed computing service supply contracts worth about 2 trillion won with an AI inference company, and related revenue will be generated when the service launches in the first half of next year,” Choi said.

What changes at launch?

NAVER has completed third-party allotment share offerings to both Brookfield Computing and Nvidia. The share allotments minimize the initial capital burden and establish a structure capable of executing the project. Choi also said early customer acquisition is proceeding smoothly.

The signed work gives the business an initial customer before service starts, but NAVER does not plan to rely on that account alone. It will pursue medium- and long-term agreements with large clients while adding global AI companies with high growth potential. Management targets steadier revenue, stronger profitability and less dependence on any single customer.

Why could the first customer generate more work?

AI models expand in scale alongside computing demand, making Choi confident the secured inference customer will lead to follow-on contracts. NAVER plans to center additional agreements on high-growth companies and keep increasing supply volume.

The contracts have also brought major investors’ recognition of the business’s growth potential, according to Choi. For NAVER, the next commercial test is whether the initial contract expands into a multi-customer portfolio.

How is NAVER positioning Vera Rubin supply?

NAVER identifies on-schedule access to next-generation GPUs as a core competitive advantage. Choi said competition to secure Nvidia’s Vera Rubin intensifies, but few sites have enough power and liquid cooling for the platform.

“Competition to lock up Nvidia’s next-generation GPU, Vera Rubin, is intensifying, but there are only a limited number of sites with enough power and liquid-cooled data centers to accommodate them. NAVER has secured this infrastructure and holds a competitive advantage in being able to supply the latest GPUs that global companies want, on time,” Choi said.

The company will use large-scale GPU orders to secure next-generation platforms ahead of rivals and respond to customer demand. On procurement, NAVER is pursuing Asia’s largest order, which lays the groundwork for offering the Vera Rubin platform before competitors.

NAVER’s current client discussions include developers of large AI models. Choi said the company is pursuing short-term and medium- to long-term contracts with those clients and holding talks with multiple other customers. It is also negotiating prices for the adoption of the latest GPUs.

What is the commercialization advantage?

NAVER says it can shorten the interval between building infrastructure and delivering computing services. Once GPU equipment arrives, the company plans to commercialize quickly, bring monetization forward and secure high utilization and profitability.

Choi expects the AI factory business to generate meaningful revenue and results. For practitioners tracking AI infrastructure, the immediate markers are the first-half launch, conversion of the initial inference contract into follow-on work, and whether NAVER turns Vera Rubin supply into contracted capacity at the prices it negotiates.