www.ptreview.co.uk
11
'26
Written on Modified on
NVIDIA and LG Group Partner to Build Modular AI Factories
NVIDIA and LG Group collaborate to deploy an integrated AI factory platform spanning robotics, autonomous driving, data center infrastructure, and sovereign AI models.
www.nvidia.com
NVIDIA and LG Group have entered a comprehensive partnership to construct an accelerated computing AI factory. The infrastructure will provide LG Group with the specialized hardware and software environment needed to train, simulate, validate, and deploy physical artificial intelligence and automation applications across its core business units.
Context of the Cooperation
Industrial manufacturing, logistics, and automotive sectors face complex operational challenges as they transition toward autonomous systems. Simulating robotic behavior in physically accurate environments, generating high-quality training data, and overcoming the severe power and thermal demands of next-generation liquid-cooled data centers require a significant level of technical integration.
To scale these operations without disrupting existing workflows, a multi-company approach was required. The collaboration combines NVIDIA’s full-stack, end-to-end AI factory platform and accelerated computing architecture with LG Group’s global manufacturing data, consumer electronics expertise, telecommunications network, and component portfolios. This joint effort enables the entities to connect AI model development, digital twins, and edge deployment into a unified workflow.
Technical Solution and Responsibilities
The technological core of the partnership relies on connecting NVIDIA's hardware and software ecosystems with LG's individual business affiliates, which are divided into distinct functional areas:
- Robotics and Physical AI: LG Electronics is integrating the NVIDIA Isaac Sim and NVIDIA Isaac Lab open robotics frameworks into its development workflows to simulate and train home-based robots, such as CLoiD, in virtual environments. LG is also exploring the NVIDIA Isaac GR00T vision-action language model for humanlike reasoning. To address data scarcity, LG Electronics is using NVIDIA Cosmos world foundation models for synthetic data generation within a physical AI data factory. Concurrently, LG Innotek is developing optical sensing components optimized for NVIDIA’s GPU architecture, while LG CNS is incorporating these systems into its PhysicalWorks industrial robot platform.
- Infrastructure and Thermal Management: The companies are developing modular AI factories aligned with the NVIDIA DSX platform. LG Electronics is driving technical collaboration on prefabricated modular designs and liquid-cooling solutions, including cooling distribution units (CDUs) and cold plates. For the underlying power and network infrastructure, LG Uplus, LG Electronics, and LG Energy Solution are building power-efficient data centers to house NVIDIA GPUs, with LG Energy Solution designing emerging 800-volt direct-current data center energy solutions in alignment with NVIDIA’s Battery Energy Storage System (BESS) guidelines.
- Autonomous Driving: LG Electronics is aligning its advanced driver-assistance systems (ADAS) and in-vehicle cockpits with the NVIDIA DRIVE Hyperion architecture and DRIVE AGX accelerated compute platform. LG Innotek will supply complementary sensing, connectivity, and lighting components engineered for this architecture.
- Sovereign AI Development: LG AI Research is leveraging NVIDIA Blackwell GPUs, the NVIDIA NeMo framework, and NVIDIA Nemotron open datasets to develop and advance EXAONE, Korea’s open sovereign AI model family, while using NVIDIA TensorRT-LLM software to build optimized inference engines for its ChatEXAONE enterprise chatbot.
Deployment or Implementation
The implementation strategy spans multiple deployment phases across global manufacturing, computing, and logistics sites. Initial development focuses on testing and validating the prefabricated modular data centers and liquid-cooling infrastructure to handle the power requirements of next-generation GPU deployments.
The physical AI components and robotic workflows will be deployed across LG CNS logistics and manufacturing floors to establish a new global smart factory standard. This autonomous manufacturing ecosystem will link the entire supply chain—from raw material procurement to product delivery—in real time through data. Concurrently, the automotive and sovereign AI solutions are being integrated into future product roadmaps for international consumer and enterprise markets.
Edited by Romila DSilva, Induportals Editor, with AI assistance.
The implementation strategy spans multiple deployment phases across global manufacturing, computing, and logistics sites. Initial development focuses on testing and validating the prefabricated modular data centers and liquid-cooling infrastructure to handle the power requirements of next-generation GPU deployments.
The physical AI components and robotic workflows will be deployed across LG CNS logistics and manufacturing floors to establish a new global smart factory standard. This autonomous manufacturing ecosystem will link the entire supply chain—from raw material procurement to product delivery—in real time through data. Concurrently, the automotive and sovereign AI solutions are being integrated into future product roadmaps for international consumer and enterprise markets.
Edited by Romila DSilva, Induportals Editor, with AI assistance.
