User Guide

Jetson AGX™ Orin Developer Kit Reviewer's Guide
8
Performance on Vision AI and Conversational AI Models
Table 3 Jetson AGX Orin Performance on widely used Vision and Conversational AI models
Faster Development with the NVIDIA AI
Software Platform
The class-leading performance and energy efficiency of Jetson AGX Orin is backed by the same
powerful NVIDIA AI platform that is deployed in GPU-accelerated data centers, hyperscale servers, and
powerful AI workstations. The NVIDIA AI platform brings optimized AI tools, libraries, and NVIDIA
SDKs such as CUDA, CuDNN, TensorRT, DeepStream, RIVA, TAO and Isaac to the Jetson platform,
enabling developers to seamlessly train AI applications on powerful cloud GPUs and deploy trained
networks on Jetson-powered AI edge devices.
Tools like NVIDIA Omniverse Replicator for synthetic data generation help in creating high quality
datasets to boost model training, NVIDIA Train-Adapt-Optimize (TAO) and Pre-Trained Models
(PTM) cuts down development time by up to 10X and provides an easier and faster way to accelerate
Instructions for running the above benchmarks on Jetson AGX Orin and Jetson AGX Xavier are available
in Running Inference Benchmarks section of Appendix