NVIDIA accelerates its layout in physical AI: deepening cooperation with Hyundai Motor to promote the industrialization of robotics into a substantial implementation phase

According to Zhitong Finance APP, NVIDIA (NVDA.US) and Hyundai Motor have agreed to deepen their cooperation to transform so-called physical artificial intelligence and robotics technology into real industrial products. NVIDIA CEO Jensen Huang and Hyundai Motor Executive Chairman Chung Eui-sun outlined the expanded vision for cooperation in the field of artificial intelligence after talks in Seoul on Monday, covering mobility, manufacturing, and robotics.

NVIDIA officially accepted Hyundai's proposal to establish a data center in the Saemangeum area of South Korea, referring to it as Korea's "AI Valley." Thus, a vast landscape spanning robotics manufacturing, AI computing infrastructure, and the hydrogen energy industry is moving from blueprint to substantial implementation.

Building the Physical AI Industry Chain

During the meeting at Hyundai Motor Group's headquarters in Seoul, Huang and Chung not only clarified the roadmap for upgrading cooperation in mobility, manufacturing, and robotics but also established a specific implementation path to push robotics technology "from laboratory research to factory production lines."

The industrialization of robotics has always been a core pillar of their cooperation. In January 2026, at the CES Consumer Electronics Show, the mass production version of the humanoid robot Atlas, developed by Boston Dynamics under Hyundai, will shake the entire industry. This humanoid robot, with 56 degrees of freedom and capable of lifting 50 kilograms, represents the highest level of industrial robots globally. Boston Dynamics announced at that time that Atlas would officially go into production in 2026 at Hyundai's Robotics Metaplant Application Center (RMAC) and Google DeepMind. It is reported that Atlas will undertake a series of core manufacturing tasks such as parts sorting and assembly line quality inspection.

"Hyundai Motor excels in manufacturing, mobility, heavy industry, and large-scale production. No company is better positioned than Hyundai Motor to seize these opportunities and create this future," Huang told the media, stating that the two companies are "very, very close" to promoting the industrialization of robotics.

Huang's strategy of naming the Saemangeum project "AI Valley" is a step-by-step approach. This vast special development zone, located on the west coast of South Korea, is becoming the core of Hyundai Motor's $6 billion investment plan. According to public information, Hyundai Motor plans to invest 90 trillion won (approximately $59 billion) in Saemangeum to create a future industrial cluster that integrates AI data centers, robotics manufacturing clusters, and hydrogen plants. Prior to this, sources cited by the Korea Economic Daily had revealed that Hyundai Motor, in collaboration with the South Korean government, is in the final negotiation stage with NVIDIA regarding the establishment of an AI R&D center, with Saemangeum leading in the site selection competition.

During the meeting, Chung introduced Huang to the Saemangeum project covering AI, robotics, and hydrogen technology, and proposed that NVIDIA join hands to create AI, robotics, and data center systems. Huang jokingly remarked, "As long as there is 'delicious grilled pork belly,' I will invest." Chung pointed out that if NVIDIA officially joins the project, the two companies could create a "perfect AI ecosystem," including a joint data center Before this, Jensen Huang was invited to visit the newly renovated headquarters lobby of Hyundai Motor, which has been designed as a "Physical AI Testbed." Various robots are responsible for security patrols, cargo delivery, and even watering plants. From the high-tech interior of this building to the large-scale production capacity cluster of the New Wan Jin project, Hyundai Motor is creating a large-scale, scalable practical model for NVIDIA's "Physical AI" strategy.

The New Wan Jin project is by no means an empty slogan. According to an agreement signed in February between Hyundai Motor Group and the South Korean government, this investment of approximately 90 trillion Korean won covers multiple cutting-edge technology fields: artificial intelligence data centers, robot manufacturing factories, and hydrogen production and application facilities.

The project's goals are clear and ambitious—transforming New Wan Jin into a global hub for "Physical AI" technology innovation and manufacturing in South Korea. As the technology route moves from the laboratory to large-scale application, Hyundai Motor announced as early as October 2025 that it would procure up to 50,000 NVIDIA AI chips to build AI factories, enhance in-vehicle intelligence, autonomous driving, smart manufacturing, and robotics research and development capabilities.

What is even more noteworthy is the pace of capacity release and the accompanying industrial effects. If the New Wan Jin AI research and development center is successfully established, construction is expected to start in the second half of 2026 and be operational by 2027. By then, the supply of automotive-grade chips from NVIDIA in Korea is expected to increase significantly, directly impacting not only the technological progress of Hyundai Motor itself but also driving a large number of local Tier 2 component suppliers in South Korea into the global AI industry chain. NVIDIA has also officially launched large-scale recruitment for its South Korea research center, focusing on cutting-edge fields such as physical AI, industrial robots, digital twins, and smart hardware.

Industry observers point out that NVIDIA's choice of New Wan Jin rather than the suburbs of Seoul as the location for its research and development center reflects two strategic intentions: first, unlike Tesla's Optimus layout in North America, NVIDIA's alliance needs to rely on Hyundai Motor's manufacturing infrastructure and global robot ecosystem to create a "robot contract manufacturing" base covering Asia and the global supply chain. Second, by deeply binding with South Korean tech giants like Hyundai Motor and LG Electronics, NVIDIA aims to stay ahead of Tesla in the global physical AI race.

Year of Mass Production: 2026 as the Year of Humanoid Robot Industry Explosion

The collaboration between Hyundai and NVIDIA is not the only "robot catalyst" in the capital market; it is merely a microcosm of the wave viewed as the "year of mass production for humanoid robots" in 2026.

According to the latest data from IDC, global shipments of humanoid robots are expected to exceed 50,000 units in 2026, a year-on-year increase of 178%. Following the first breakthrough of 12,000 units in 2025, the industry is entering a true window for accelerated scaling. With operational delays reduced to below 50 milliseconds, operational precision breaking through 0.02 millimeters, and multiple iterations of dexterous hand technology, humanoid robots have evolved from "able to walk" to "able to use" in terms of productivity.

In the United States, Tesla's Optimus Gen-3 officially began mass production in the second quarter of this year at the Fremont factory, with the original Model S/X production line completely converted to robots, achieving an annual design capacity of one million units. Looking ahead to 2027 and 2028, the annual shipment volume of humanoid robots is expected to grow to hundreds of thousands or even millions, with automotive manufacturing, 3C electronics, and power inspection becoming the fastest-starting "essential needs." Application Scenarios. Jensen Huang predicts that this transformation will unlock "trillions of dollars in economic opportunities."

What sets Hyundai Motor apart in this competition is its three parallel advantages: mastering world-class humanoid robot manufacturing capabilities (Atlas program) through Boston Dynamics; becoming the core "application hub" of the robotics industry through its vast global automotive manufacturing system (parts assembly, complete vehicle assembly lines); and obtaining foundational chip and algorithm support for physical AI through deep integration with NVIDIA. This "iron triangle" model of "manufacturing assets + AI computing power + application scenarios" is currently nearly impossible to replicate among global automakers.

In contrast, Tesla's Optimus has a core advantage in extreme vertical integration, with its computing brain, neural network architecture, and key actuators all developed in-house. The Hyundai-NVIDIA alliance brings together Hyundai Motor (Boston Dynamics), BMW, and other leading automakers to develop diverse robotic forms for different application scenarios, accelerating industry implementation. Which of the two models will ultimately dominate the physical AI market will depend on the speed of large-scale production cost reduction and the richness of the application ecosystem over the next three years.

The Era of Physical AI Begins: The Growth Engine of a Trillion-Dollar Market

If the core of the AI wave over the past two years has been "cloud computing power" and the "big model parameter competition," then the strongest signal released by Jensen Huang during his visit to South Korea is the formal shift of the industry focus towards "physical AI"—the deep integration of AI technology with physical hardware, industrial scenarios, and terminal devices.

Physical AI is breaking down the barriers between AI and physical manufacturing, allowing decision-making capabilities driven by large models to "grow limbs" and "operational arms." Intelligent manufacturing, automotive assembly, hazardous environment operations, and even home services will constitute a massive market far exceeding traditional data center investment scales. Multiple industry research reports, including those from Huayuan Securities, indicate that the physical AI industry is expected to open up a new trillion-dollar market.

Several strategists believe that the physical AI track is entering an industrial validation period of "moving from concept to capacity investment," and investors need to shift their focus from purely AI chip narratives to hardware-level beneficiaries such as complete robots, intelligent sensors, and edge computing modules