Nvidia Expands Into Robotics and Physical AI Through LG and Foxglove Collaborations

Nvidia (NasdaqGS: NVDA) is expanding its presence in robotics and physical AI through new collaborations with LG Electronics and Foxglove. The company is working on an accelerated robotics Data Factory with LG, which combines real-world and synthetic data for robot learning.

Foxglove has launched agentic AI tools that use NVIDIA's Cosmos world models to power semantic robotics data search and debugging workflows.

These moves demonstrate Nvidia's efforts to extend its role beyond AI chips and cloud infrastructure into commercial robotics data platforms and physical AI ecosystems.

This development aligns with the company's narrative of turning its AI stack into full infrastructure for physical AI, not just data center compute. LG and Foxglove's collaborations lean on NVIDIA Cosmos and Isaac, which deepens customer reliance and increases switching costs.

However, this also touches on a key risk around rising data center and power constraints, as scaling physical AI and robotics workloads depends on energy and infrastructure capacity that is already a pressure point. The news tilts modestly toward the bull case by showing real adoption of NVIDIA's world and robotics models.