NVIDIA and LG Group are expanding their partnership to build an AI factory aimed at accelerating the Korean conglomerate’s development of robotics, autonomous driving, data centre technologies and GPU cloud services.
The planned infrastructure will provide accelerated computing resources for training, simulation, validation and deployment of AI applications across LG’s businesses. The collaboration combines NVIDIA’s full-stack AI factory platform with LG’s capabilities in consumer electronics, robotics, automotive components, smart spaces and data centre infrastructure.
At the centre of the initiative is an effort to connect AI model development, physical AI data generation, robot simulation, training, edge deployment and factory-scale digital twins into a common development environment.
LG is expected to use NVIDIA’s robotics technologies to expand its work on physical AI, including home and industrial robots. LG Electronics is developing home robots such as CLOiD, while its engineering teams are exploring NVIDIA Isaac Sim and Isaac Lab for simulation, training and validation. The platforms allow robotic systems to be tested in virtual environments before deployment in physical settings.
LG is also evaluating NVIDIA Isaac GR00T, a vision-language-action model designed for humanoid and general-purpose robotics. The companies plan to jointly develop reference robots, potentially integrating LG’s systems into the broader Isaac GR00T ecosystem.
A major focus will be generating the large volumes of training data required for physical AI. LG Electronics plans to develop a physical AI data factory using NVIDIA Cosmos world foundation models to generate and augment synthetic data for robotics and industrial AI applications.
LG Innotek will contribute sensing and optical technologies for robotics, while LG CNS plans to incorporate NVIDIA Isaac, Cosmos and GR00T technologies into its PhysicalWorks platform for manufacturing and logistics applications.
The companies also intend to combine LG’s production data and manufacturing expertise with NVIDIA’s digital-twin technologies to develop more autonomous manufacturing environments. The objective is to connect procurement, production, logistics and delivery through real-time data and AI systems.
The partnership also covers the physical infrastructure required to support increasingly demanding AI workloads. LG Electronics is working with NVIDIA on cooling technologies for AI factories, including cooling distribution units and cold plates. The companies are also collaborating on prefabricated modular designs intended to accelerate the deployment of liquid-cooled AI infrastructure.
The work is aligned with NVIDIA’s DSX AI factory platform, which is designed to support scalable accelerated computing environments. Within LG Group, LG CNS plans to develop high-performance AI factories using NVIDIA GPUs and DSX technologies. LG Uplus is preparing to build a large-scale AI data centre capable of accommodating NVIDIA’s latest GPUs, while also exploring GPU cloud and AI service opportunities.
LG Energy Solution is collaborating with NVIDIA on 800-volt direct-current power technologies for data centres. The work is intended to address the growing power requirements of next-generation accelerated computing systems.
The collaboration extends into automotive technology, where LG Electronics is aligning its advanced driver-assistance and in-vehicle AI systems with NVIDIA’s DRIVE platform. LG plans to align sensor, computing and software architectures with NVIDIA DRIVE Hyperion while exploring NVIDIA DRIVE AGX for future mobility applications. Areas of development include AI-powered cockpits and edge AI processing for software-defined vehicles.
LG Innotek will also work with NVIDIA on sensing, connectivity and lighting components designed around NVIDIA’s automotive computing architecture.
The partnership also encompasses LG AI Research’s EXAONE family of AI models. LG has used NVIDIA Blackwell GPUs, the NeMo framework, Nemotron open datasets and TensorRT-LLM software in developing and deploying EXAONE.
LG Group is exploring wider use of EXAONE and agentic AI across its businesses, including through ChatEXAONE, its enterprise chatbot platform. NVIDIA’s infrastructure and software are expected to support further development of LG’s sovereign AI capabilities, while LG aims to apply the technology across enterprise operations.
The expanded partnership spans the AI stack from computing infrastructure and data centre technologies to model development, robotics, industrial automation and vehicles. For NVIDIA, the collaboration extends its AI factory and physical AI technologies into a major global industrial group. For LG, it provides a common computing and software foundation for deploying AI across businesses increasingly dependent on real-time data, simulation and accelerated computing.
