Nvidia finalizes payment of US$1 bln, acquires over 4 pct stake in Naver

Nvidia acquired a partial ownership stake in Naver, South Korea’s portal operator. It did not acquire the whole company. The stake represented 4.13 percent of Naver’s shares after the transaction. Nvidia paid US$1 billion for that investment.

Naver issued more than 6.5 million new shares through a private placement involving Nvidia. The company said it received 1.3 trillion won, equal to about US$1 billion, in proceeds. This means the money went to Naver as investment capital rather than simply changing hands between existing shareholders.

The purchase matters because it is tied to the companies’ joint artificial intelligence factory venture. Naver plans to build GPU-based infrastructure, beginning at its Gak Sejong data center. The article says Nvidia’s stake acquisition is set to speed up that project and Naver’s planned GPU-as-a-service expansion.

A 4.13 percent stake means Nvidia owns shares representing 4.13 percent of Naver’s issued share ownership. Put simply, if Naver had 100 equal shares, Nvidia’s portion would be a little more than four shares. The article identifies this as Nvidia’s new stake after the investment was completed.

This percentage gives Nvidia a financial ownership interest in Naver. Nvidia can benefit if the value of its Naver shares rises and is exposed to the investment’s risks. However, the percentage alone does not mean Nvidia owns Naver or controls all of its decisions. The article does not describe voting rights or board representation.

The stake also connects the companies more closely. Nvidia’s investment is expected to accelerate their joint artificial intelligence factory venture. Naver plans GPU-based infrastructure and aims to expand its GPU-as-a-service business, starting with its Gak Sejong data center.

A private placement is a share sale arranged directly with selected investors. A third-party rights offering similarly gives a specific outside investor the opportunity to subscribe to newly issued shares. Unlike a broad public offering, the company negotiates the transaction with the named investor and receives the investment proceeds directly.

In this case, Naver issued more than 6.5 million new shares through a private placement involving Nvidia. Nvidia paid US$1 billion, or 1.3 trillion won, and received 4.13 percent of Naver. Because the shares were newly issued, Naver received the money as corporate funding, while existing shareholders’ percentage ownership could be diluted.

The arrangement gives Naver capital for its plans without describing a loan or asset sale. It also gives Nvidia a direct ownership position in its Korean partner. The article links the completed payment and stake acquisition to faster progress on their joint artificial intelligence factory venture.

The AI factory is a joint venture announced by Nvidia and Naver in June. The article describes it as an effort to construct infrastructure based on graphics processing units, or GPUs. The infrastructure will support Naver’s plans to expand its GPU-as-a-service businesses.

Naver plans to begin with its Gak Sejong data center in the central city. Nvidia has now completed its US$1 billion investment and acquired 4.13 percent of Naver through newly issued shares. The article directly links that completed stake acquisition with faster progress on the joint venture.

The investment provides Naver with 1.3 trillion won in proceeds. The article does not specify a construction timetable, the factory’s final size, or how the partners will divide operations. It does establish the next direction: GPU-based infrastructure first, followed by expansion of GPU-as-a-service activities.

Naver’s Gak Sejong data center will serve as the starting location for the GPU-based infrastructure planned under its partnership with Nvidia. The article places the facility in Sejong, a central city in South Korea. It identifies the data center as the first step rather than describing the entire project’s final footprint.

The key mechanism is infrastructure built around GPUs. These processors provide the computing capacity Naver intends to use for its GPU-as-a-service business. In practical terms, the data center is the physical base from which Naver can develop and offer that service, although the article gives no detailed technical design.

Nvidia’s completed US$1 billion investment is expected to accelerate the joint AI factory venture. That makes Gak Sejong an important early implementation point. The article does not state when construction will finish, how many GPUs will be installed, or how much service capacity the center will provide.

A GPU, or graphics processing unit, is a processor that can perform many similar calculations in parallel. This differs from relying mainly on a processor that handles fewer tasks in sequence. GPUs were developed for graphics, but their parallel structure also fits many artificial intelligence workloads, which involve processing large amounts of data and mathematical operations.

For example, training or running an AI model can require repeated calculations across numerous data points. A GPU can work on many of those calculations simultaneously. That can make AI computing faster and more practical, especially when many processors are combined in data-center systems. These technical details come from established computing knowledge, not from the article itself.

The article says Naver plans GPU-based infrastructure beginning at Gak Sejong. It does not give specifications or explain performance targets. Nvidia’s investment is expected to speed the joint AI factory venture, which will support Naver’s GPU-as-a-service plans.

GPU-as-a-service is a model in which a provider makes GPU computing available to customers over a service platform. Customers can use processing capacity when needed instead of purchasing an entire GPU system. The article says Naver plans to expand this business after building GPU-based infrastructure.

For example, a company could request GPU capacity for an AI workload and pay for access according to the provider’s terms. The provider operates the underlying infrastructure, while the customer uses the computing resource. This can help companies avoid buying and maintaining their own hardware, especially when demand changes. These general operating details are established industry knowledge, not specifics supplied by the article.

Naver plans to start constructing the infrastructure at its Gak Sejong data center. Nvidia’s completed US$1 billion investment is expected to speed the joint AI factory venture. The article does not state the service’s prices, customers, capacity, or launch date.