Reduced Financial Risk, but Profitability Remains Uncertain

Break Free from the Burden of Massive Fixed Costs

Competing with Global Tech Giants... Securing Customers Comes First

[The View] Naver AI Factory: Utilization and Demand Matter More Than Investment Size View original image

Recently, Naver announced plans to build a large-scale artificial intelligence (AI) factory centered in Sejong. In simple terms, an AI factory is a massive computing facility that brings together all the servers, power supply, and cooling systems required to train and operate AI models. Just as a conventional factory produces goods using machinery and equipment, an AI factory develops AI models and services using vast computing resources. The total project is valued at $10 billion. Brookfield, a major financial player, will provide $9 billion in funding, while Nvidia is investing $1 billion in Naver. The Ministry of Science and ICT and the Office of the President introduced this as a concrete achievement of AI collaboration during official briefings.


However, as the AI bubble continues to grow, it is crucial to soberly assess whether this investment can generate profits, rather than simply celebrating its scale.


One notable aspect of this project is that Naver is not directly purchasing all the equipment, such as GPUs and network infrastructure. Under the current plan, Brookfield is participating as the capital and infrastructure partner, while Naver will lead the operation of the AI factory. As the operator, Naver will pay usage fees for the resources it uses, and plans to generate revenue by selling GPU computing resources and related assets to corporate clients. Because Naver does not bear the full upfront capital investment, this mitigates financial risk. However, sharing the financial burden is quite different from actually realizing profits from the venture.


First, it is unlikely that Brookfield, which is raising $9 billion, would assume all the risks of building the infrastructure without any revenue guarantees. The final agreement will almost certainly include mechanisms to protect the investor's returns. There are also risks inherent in the business model itself. While the idea of selling AI computing resources to external clients may be feasible, securing customers who can pay for these resources reliably over the long term is a different challenge. It remains uncertain whether Naver can offer a competitive advantage over global cloud providers such as Amazon, Microsoft, and Google, as well as other AI data center operators, and whether it can secure enough ongoing customers amid fierce competition. If the number of external customers is lower than expected, the data center's utilization rate will decline. In that case, Naver could fail to recover its operating expenses, as well as the fees owed to Brookfield.


Naver's new venture should not be judged solely by the size of investment or the number of GPUs secured. What matters most is the identification of confirmed customers. The utilization rate and profitability of the center, based on this scale, must be evaluated objectively. At the same time, it is vital to clarify exactly what competitive advantages Naver's AI factory offers compared to established global players. In the fiercely competitive AI data center sector, the risk profile differs significantly between building out infrastructure first and then seeking customers, and securing customers first before expanding infrastructure to meet demand.


Just as building a highway is meaningless if there are no cars to drive on it, acquiring an AI factory and GPUs does not automatically guarantee business success or establish leadership in AI. Only when there are services and companies that consistently use the infrastructure and are willing to pay for it does the AI factory become a productive asset. Without this, a data center filled with cutting-edge GPUs could become nothing more than a massive warehouse racking up enormous fixed costs.



Kyungnakyung, Professor of Business Administration, Sungkyunkwan University


This content was produced with the assistance of AI translation services.

© The Asia Business Daily. All rights reserved. Unauthorized AI training and use prohibited.

Today’s Briefing