NVIDIA Partners With LG To Build AI Factory For Robotics

Major technology company NVIDIA has partnered with South Korean conglomerate LG Group to create an AI factory.

The new facility gives LG Group the ability to train, simulate, validate and deploy AI-centred applications through its various businesses which include electronics, house appliances, and more.

The companies say they will start the process by building an autonomous manufacturing ecosystem to spearhead the operation.

NVIDIA will provide end-to-end AI infrastructure while LG’s expertise in production technology and global manufacturing allows the factory to function efficiently.

Each step of the process, from raw material procurement to production, logistics and customer delivery, will be supported by both data and AI.

Robotics and AI framework

LG Electronics is developing robots, like CLoiD, which can do everyday indoor household chores and duties. The robots will integrate with NVIDIA’s Issac Sam and Issac Lab robotic frameworks, allowing LG to simulate, train and validate them in virtual environments prior to arrival.

NVIDIA is also using its Isaac Groot open reasoning vision action language model to provide the LG robots with humanlike reasoning the ability to complete complex tasks. Both corporations will also create reference robots for further integration.

A physical AI data factory is also being built to compensate for any training data challenges that may arise within robotics. The factory will provide high-quality training data for any robotics or industrial AI project and use NVIDIA Cosmos world foundation models for synthetic data generation and augmentation.

While LG innotek handles robotic components, particularly sensing solutions, LG CNS will create an ecosystem allowing other corporations to seamlessly incorporate AI robots within manufacturing and logistics.

Thermal dynamics and autonomous driving

NVIDIA is providing cooling solutions for AI factory thermal management and LG Electronics uses its technology for CDUs, cold plates and prefab modular design capabilities to tackle power, thermal and deployment challenges within future liquid-cooled AI factories.

Future factories will be built from NVIDIA’s DSX infrastructure in collaboration with LG Electronics, LG Energy Solution and LG Uplus. LG Energy Solution also plans to collaborate with NVIDIA to build 800 volt-direct-current data centre energy solutions.

Through this partnership, NVIDIA and LG AI Research will also advance EXAONE. This is a leading AI model in Korea backed by NVIDIA’s NeMo framework and one that’s available to developers, enterprises and researchers. 

The announcement follows NVIDIA’s launch of Cosmos 3, which is an open world model built for physical AI, combining vision reasoning, world generation and action prediction.

Jensen Huang, Founder and CEO of NVIDIA, says: “The big bang of physical AI is just around the corner thanks to breakthroughs in multimodal reasoning language, vision and world models. 

“The Cosmos 3 family of open, frontier omnimodels gives developers a generational leap in ability to build robots, AVs and vision AI that perceive, reason, plan and act in the physical world.”