NVIDIA and LG Group partner to build an AI factory spanning robotics, ADAS, and sovereign AI

NVIDIA and LG Group announced on June 7, 2026 a broad collaboration to build an AI factory designed to accelerate LG's businesses across robotics, autonomous driving, data center infrastructure, and sovereign AI.

What's new

"The AI factory will provide LG Group with accelerated computing infrastructure to train, simulate, validate and deploy AI-based applications across its key businesses," NVIDIA stated.

The partnership spans four LG subsidiaries and five distinct technical domains:

Robotics (LG Electronics): LG will deploy NVIDIA Isaac Sim, Isaac Lab, and Isaac GR00T to develop home robots including CLoiD, LG's consumer robot platform. A physical AI data factory will generate robot training data using NVIDIA Cosmos world-foundation models.

Autonomous driving (LG Electronics): LG ADAS and in-vehicle systems will align with the NVIDIA DRIVE Hyperion architecture and DRIVE AGX compute platform.

Data center infrastructure (LG CNS, LG Electronics): Implementation of NVIDIA DSX-aligned infrastructure with liquid cooling and modular design for scalable data center buildout.

GPU cloud services (LG Uplus): LG's telecommunications arm will offer GPU cloud services powered by NVIDIA hardware.

Sovereign AI (LG AI Research + NVIDIA): Joint work on EXAONE — described as Korea's leading sovereign AI model — using NVIDIA Blackwell GPUs, the NeMo framework, and TensorRT-LLM for deployment.

Context

This is the second major NVIDIA plus Korean industrial conglomerate AI factory announcement in two days. On June 7, NVIDIA also announced a separate expanded collaboration with Doosan Group covering physical AI, robotics, and AI factory infrastructure. Both announcements reflect an aggressive push to embed NVIDIA's full AI stack into Korean conglomerates.

LG Group is one of Korea's largest chaebol, with significant operations across consumer electronics (LG Electronics), energy storage (LG Energy Solution), chemicals, and telecommunications (LG Uplus). EXAONE is LG AI Research's large language model, one of the few corporate-driven sovereign AI initiatives in Asia operating independently of state funding.

Why it matters

The breadth of the partnership is the defining characteristic: robotics, ADAS, cloud infrastructure, and sovereign AI in a single agreement. That scope suggests LG Group is not piloting AI in one division but making a systemic platform bet on NVIDIA across its entire business.

The sovereign AI dimension — EXAONE trained on Blackwell GPUs using NeMo — is strategically distinctive. As enterprises and countries seek AI capabilities not dependent on US-hosted models or API access, the combination of locally-trained models with leading-edge hardware represents an emerging infrastructure pattern. LG and NVIDIA are effectively co-building a Korean AI sovereignty stack at scale.

For NVIDIA, the back-to-back Korean conglomerate announcements reinforce a strategy that goes well beyond selling GPUs to US hyperscalers. Industrial AI factories — spanning simulation, training, and deployment across manufacturing, ADAS, and robotics — represent a growing NVIDIA revenue segment, and Korean chaebols represent high-volume, long-duration infrastructure customers.