NVIDIA has recently announced Project GR00T, which marks a significant leap in the field of humanoid robotics. Standing for Generalist Robot 00 Technology, GR00T is a general-purpose foundation designed to give humanoid robots the ability to process natural language and replicate human motions. During the GTC keynote, NVIDIA's CEO, Jensen Huang, showcased robots powered by GR00T performing various tasks with remarkable dexterity and coordination. These robots are anticipated to have the capability to learn, adapt, and interact within human environments swiftly. A critical component of this initiative is the introduction of Jetson Thor, a robot computer built on the NVIDIA Thor system-on-a-chip (SoC). Jetson Thor exemplifies a modular approach to computing, engineered to carry out intricate tasks while maintaining a harmonious and safe interaction with humans and machines. Featuring a next-gen GPU, the SoC delivers extraordinary 8-bit floating-point AI performance to run multimodal generative AI models such as GR00T. Furthermore, with a safety processor and vast ethernet bandwidth, it streamlines design and integration processes widely. The upgrades to the NVIDIA Isaac™ robotics platform cannot be understated, incorporating tools vital for simulation and generative AI foundation models, critical for the development of humanoid robots. This involves the new Isaac Lab, enhancing robot learning through parallel simulations, and OSMO, a service orchestrating compute workflows across diverse environments. Additionally, the launch of Isaac Manipulator and Isaac Perceptor expands NVIDIA's offerings with a suite of pre-trained models and hardware, aiming to proliferate robotic arms' dexterity and multi-camera vision systems within industries focused on operational efficacy and safety.