Here I rest my notes on every usefull thing I learned about the jetson, but In simple language and terms.

Why is Jetson Xavier AGX usefull?

Combined with his high processing power and GPU, optimized to fully use cuda and run machine learning models very quickly. Jetson xavier Agx is a good solution for vision based problems who require high computing power.

Jetson has some very usefull things:

• Jetpack (Nvidia’s Software and Go-To Kit) has cuda, tensorrt , deepstream cudaAI installed. (This saves a lot of hassle configuring and deploying. Installing this on a normal computer could proove to become a big pain)
• DeepstreamSDK (Quick deployment of Optimized DL models on video streams)
• TOA Toolkit (Transfer learning toolkit, makes training Neural Networks much quicker)
• Jetson Inference (NVIDIA’s go to solutions optimized and easy to deploy and costumize)

NVIDIA JetPack SDK supports both your developer kit and host development platform.

It includes:  Sample Linux filesystem with NVIDIA drivers  AI and Computer Vision libraries and APIs  Developer tools  Supporting documentation  Code samples to help you get started

JetPack Components OS Image JetPack includes a sample file system derived from Ubuntu.

Libraries JetPack includes the following libraries:

 TensorRT and cuDNN for high performance deep learning applications  CUDA for GPU accelerated applications across multiple domains  Multimedia API package for camera applications and sensor driver development  VisionWorks and OpenCV for visual computing applications