NVIDIA and Doosan Group Partner on Physical AI and Data Centers
The collaboration spans robotics, power infrastructure, and advanced materials for next-generation AI factory deployments.
NVIDIA is deepening its collaboration with South Korean industrial conglomerate Doosan Group to advance physical AI capabilities and the infrastructure needed to power next-generation AI factories.
The partnership brings together NVIDIA's accelerated computing platforms with Doosan's expertise in industrial automation, power generation, and electronics materials. The collaboration touches multiple Doosan business units including Doosan Robotics, Doosan Bobcat, Doosan Enerbility, and Doosan Corporation Electro-Materials.
Why it matters
As AI systems move from cloud-based inference to physical robots and autonomous equipment, companies need both the software stack and the industrial infrastructure to deploy them at scale. This partnership addresses the full vertical—from the materials inside data center servers to the power plants that run them to the robots they enable. For enterprises investing in physical AI, it signals how traditional industrial players are repositioning around the technology.
Robotics and autonomous equipment
Doosan Robotics is integrating several NVIDIA technologies into its Agentic Robot OS platform, including Isaac Sim and Isaac Lab robotics frameworks, Cosmos world foundation models, the Newton physics engine, and Jetson Thor edge computing hardware. The goal is to help industrial robots better perceive their environments, reason about tasks, and act in dynamic settings.
The companies plan to develop reference implementations for high-value industrial applications such as depalletizing and sanding operations. They're also exploring new robot form factors including dual-arm systems and humanoid platforms.
Doosan Bobcat, which manufactures construction and landscaping equipment, plans to integrate NVIDIA's physical AI technologies into compact machinery used across construction, agriculture, and material handling. The work aims to create specialized world models that enable equipment to operate more autonomously in diverse outdoor environments.
Power infrastructure for AI factories
Doosan Enerbility is exploring how its power generation portfolio—including gas turbines, steam turbines, small modular reactors, and hydrogen fuel cells—can support AI data centers that require continuous, high-efficiency power delivery.
Future collaboration may include power supply design for AI factory deployments, optimization of generation equipment, and evaluation of low-carbon sources such as small modular reactors. The work addresses the growing energy demands of accelerated computing infrastructure.
Advanced materials for AI servers
Doosan Corporation Electro-Materials is supplying copper clad laminate (CCL), a foundational material for printed circuit boards used in AI accelerators, networking equipment, and server motherboards. High-performance CCL enables low signal loss and high reliability in systems where bandwidth and signal integrity are critical.
The materials support NVIDIA's MGX modular reference architecture, which helps system manufacturers build servers and rack-scale AI infrastructure. As AI systems increase in performance, advanced PCB materials become essential for maintaining signal quality across data center equipment.
The collaboration details were first reported by NVIDIA in a company blog post.
This is an original analysis by the Omega editorial team. Source reporting: AI Watch.
Want systems like this working for your business?
Book a Call
