Featuring NVIDIA's latest-generation Jetson Thor module, it delivers up to 2070 TFLOPS of computing power—over 7.5 times the AI performance of the previous-generation AGX Orin™. This high-performance hardware is specifically engineered to support applications such as humanoid robotics, autonomous driving, spatial intelligence, and generative AI. 

High-speed networking capabilities enable communication with endpoints through multiple interfaces, reducing data forwarding and processing time during transmission. This significantly lowers communication latency while safeguarding data privacy. Extensive expandability offers users greater deployment flexibility.

GPUs based on the Blackwell architecture efficiently run large Transformer models, delivering powerful computing for generative AI and real-time inference. NVIDIA's full-stack software ecosystem for edge AI and robotics applications covers every stage from model training to deployment. The Holoscan platform simplifies real-time processing of high-speed sensor data. 

GR00T pre-trained models and the Isaac platform accelerate robotics development iterations. Leveraging the Omniverse platform enables precise physical simulation and iterative testing in virtual environments. Supported by a comprehensive ecosystem of software and hardware, this product emerges as an ideal platform for future embodied intelligence industry applications.