NVIDIA (NVDA.US) is forging a partnership with South Korea's LG Group in the fields of humanoid robotics and next-generation data centers. This move underscores NVIDIA's aggressive push into the realm of "Physical AI" (embodied intelligence) beyond its traditional AI chip business. Following a meeting in Seoul with LG Group Chairman Koo Kwang-mo, CEO Jensen Huang stated, "We are working with them on motor technology and mechanical systems to integrate humanoid robotics technology with the future of robots."
He added, "We are also working with LG to build the data centers of the future." NVIDIA commented, "This collaboration perfectly combines NVIDIA's full-stack, end-to-end AI factory platform with LG Group's global leadership in consumer electronics, robotics, mobility components, smart spaces, and data center technology."
Expanding into Physical AI and AI Infrastructure
NVIDIA CEO Jensen Huang visited South Korea in early June after concluding events in Taipei, securing a series of strategic agreements with several leading Korean corporations including LG Group, SK Telecom, SK hynix, Naver, and Doosan Group. This visit marks a significant, concrete step in NVIDIA's strategic shift from a traditional AI chip supplier to a provider of "Physical AI" and comprehensive AI infrastructure services.
In addition to the new partnership with LG, NVIDIA reached agreements with the following companies:
SK Telecom: Building a Gigawatt-Scale AI Cloud
SK Telecom plans to utilize NVIDIA's DSX™ platform to construct a gigawatt-scale AI cloud in South Korea, with the first AI factory expected to launch in 2027. This collaboration aims to provide robust infrastructure support for AI innovation in Korea, meeting surging domestic and global AI computing demands.
For both parties: SK Telecom gains access to NVIDIA's most advanced AI factory platform to rapidly build and scale full-stack AI infrastructure; NVIDIA secures South Korea's largest telecom operator as a long-term client, promoting the standardized deployment of the DSX platform in the Asia-Pacific region.
SK hynix: Multi-Year HBM Technology Partnership
NVIDIA and SK hynix announced a multi-year technology partnership to jointly advance next-generation memory technologies for global AI factory construction and to accelerate semiconductor design and manufacturing. Notably, during COMPUTEX, Jensen Huang wrote "Please Make More" on an HBM4E wafer at SK hynix's booth, highlighting NVIDIA's urgent need for high-bandwidth memory (HBM) supply.
SK hynix solidifies its position as a core HBM supplier for NVIDIA's AI accelerators and gains joint development support for HBM4 and subsequent products; NVIDIA ensures supply chain security for memory in its next-generation AI chips (such as Vera Rubin).
Naver: Expanding Sovereign AI Infrastructure
Korean internet giant Naver will use NVIDIA's DSX™ platform to expand its sovereign AI infrastructure, starting with an initial 55-megawatt scale and planning to gradually expand to gigawatt levels to address the global surge in AI demand. Jensen Huang plans to visit Naver's "1784" robot-friendly headquarters in Pangyo.
Naver builds autonomous, controllable AI infrastructure to support its large language models like HyperCLOVA and cloud services, reducing reliance on overseas platforms; NVIDIA establishes a benchmark case in Korea's internet and AI service sector, promoting the implementation of the "sovereign AI" concept.
Doosan Group: Physical AI and AI Factory Infrastructure
NVIDIA and Doosan Group expanded their collaboration to cover multiple subsidiaries, including Doosan Robotics, Doosan Bobcat, Doosan Enerbility, and Doosan Electronic Materials. Areas of cooperation include Physical AI and robotics technology, AI factory infrastructure, power infrastructure (to meet the surging energy demands of AI data centers), and semiconductor materials (such as CCL, copper-clad laminate).
According to Korean media reports, Doosan Electronic Materials is expected to become the exclusive supplier of CCL materials for NVIDIA's Rubin chips, with related sales projected to reach 1 trillion won in 2026.
For both parties: Doosan Robotics gains support from NVIDIA's Isaac platform and simulation technology, the energy division addresses AI data center power needs, and the electronic materials division enters NVIDIA's core supply chain; NVIDIA completes its industrial chain layout from robotics hardware to energy supply, ensuring localized support for AI factory construction.
From Selling Chips to Selling AI Factories
The series of agreements secured during Jensen Huang's visit to South Korea illustrate NVIDIA's clear strategic transformation path—from a pure GPU chip supplier to a platform company offering full-stack AI factory solutions. Through comprehensive cooperation with Korean leaders in memory (SK hynix), telecommunications (SK Telecom), internet (Naver), consumer electronics and robotics (LG), and heavy industry and energy (Doosan), NVIDIA is building a complete ecosystem covering "chips-software-infrastructure-applications."
For South Korea, these collaborations will accelerate the construction of its "K-AI" ecosystem, enhance its influence in the global AI industry, and drive the transformation of traditional industries for the AI era. This visit is seen as the implementation of substantive cooperation following the 2025 "Kangbu meeting" (where Huang met with Samsung's Lee Jae-yong and Hyundai's Chung Eui-sun), marking South Korea's emergence as NVIDIA's most crucial partner region outside the United States and Taiwan, China, within its global AI ecosystem.
Comments