LG puts an 80MW factory behind physical AI

LG and NVIDIA are moving from a strategic partnership to an 80MW AI factory in Cheonan, South Korea, targeted for the first half of 2028, alongside a humanoid unveiling next year and wheel-based robot validation on LG’s Tennessee washing-machine line. LG’s announcement says the factory will use NVIDIA’s Vera Rubin platform to develop physical-AI and robot foundation models. Builders should file the schedule as a capacity signal, not a deployment promise: the useful evidence arrives when the Tennessee pilot produces measured uptime and task economics, much as World Labs’ simulation-to-real funnel ties robotics claims to hardware trials.