Alliance with 22 Japanese Robot Firms: "NVIDIA Remains the Leader in the Physical AI Era" [Click e-Stock]

KB Securities Maintains "Overweight" Rating
"NVIDIA's Brains, Japan's Bodies Achieve Synergy"

KB Securities has maintained its "Overweight" investment rating on NVIDIA, stating that the company has successfully brought 22 leading Japanese robot manufacturing firms into its physical AI coalition, the "Cosmos Alliance." Though no separate target price was suggested, the firm noted that NVIDIA's valuation—reflecting its future sales growth—remains below the industry peer average.

On July 22, Yoo Joong-ho, analyst at KB Securities, commented, "The country that knows factories best has decided to build the brains of its robots on the NVIDIA platform," adding, "NVIDIA is expected to solidify its leadership position in physical AI, following its dominance in data centers."

In this partnership, NVIDIA will not manufacture robots directly. Instead, the company will provide the essential learning infrastructure, virtual simulation environments, and the AI models that function as the robot "brain," all of which are commonly needed by robot manufacturers.

Within this structure, NVIDIA's revenue is not tied to the success or failure of any single robot, but rather depends on how many manufacturers adopt its platform. Yoo explained, "The participation of Japanese companies, which operate in more industrial settings than any other country, suggests that this adoption is becoming an industry standard."

Jensen Huang, CEO of NVIDIA, signs a mobile robot platform called 'MobED' after meeting with Euisun Chung, Executive Chair of Hyundai Motor Group, at Hyundai Motor Company's headquarters in Seocho-gu, Seoul, on June 8. Photo by Joint Press Corps

The core product is "Cosmos3," a brain template pre-trained with simulated videos of the physical world in action. Companies can download this brain model free of charge and further train it to suit their robots and field environments. However, at each stage from training to deployment, manufacturers are required to purchase NVIDIA hardware—including data center servers, GPUs for simulation, and small modules for robot integration. Because this recurring revenue is tied to the number of robots sold, demand related to physical AI is expected to extend beyond data centers to include factories and robots as adoption grows.

The main risk identified is the potential for major manufacturers to switch to their own custom chips. If robot adoption scales up, manufacturers seeking cost reductions could develop their own inference chips instead of using NVIDIA's Jetson modules. However, given that the market is still in its early stages and rapid development is a priority, NVIDIA's standard-setting position is expected to continue for quite some time.

Yoo stated, "The addition of Japan, which boasts the largest number of industrial sites in the world, is creating a data moat that will be difficult for latecomer platforms to overcome. Factoring in its long-term sales growth potential, NVIDIA remains undervalued compared to the industry average," he added.

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