DGX Spark Live: Our Robot Adventure

In this livestream, we introduce Spark as an AI “brain server” for small mobile robots, where LLM inference runs centrally on Spark and communicates with onboard actuators over Wi‑Fi to dramatically reduce robot weight and battery usage while maintaining high AI performance. What we’ll cover: - Spark can serve as AI brain for small mobile robots: LLM inference runs on Spark as server that communicates with actuators in the robot via wifi, saving weight and battery life while giving high AI performance. - Project Mem-o: building a mobile robot with lego chassis & motors, phone as face display, raspberry pi for motion control. LLM inference on spark makes the robot intelligent, allowing it to chat, move and play small games. - Spark as AI home server - including web IDE for the Mem-o robot: a graphical programming language that allows to design new workflows