Introduction / What is Jetpack?

< >

‎

NVIDIA Jetpack Reference Guide

Table of Contents [Sticky]

• Introduction
  • What is Jetpack?
  • Why use Jetpack?
    • Benefits of Jetpack
  • Main Features
• RidgeRun Custom Development
• Version Index
  • Jetpack 4
    • Supported Platforms
    • New Features Introduced
    • Software Components and Versions
  • Jetpack 5
    • Supported Platforms
    • New Features Introduced
    • Software Components and Versions
  • Jetpack 6
    • Supported Platforms
    • New Features Introduced
    • Software Components and Versions
• Jetson Boards
  • Orin Nano
    • Characteristics
    • Supported Jetpack Versions
  • Orin NX
    • Characteristics
    • Supported Jetpack Versions
  • AGX Orin
    • Characteristics
    • Supported Jetpack Versions
  • Xavier NX
    • Characteristics
    • Supported Jetpack Versions
  • Nano
    • Characteristics
    • Supported Jetpack Versions
  • AGX Xavier
    • Characteristics
    • Supported Jetpack Versions
  • TX2
    • Characteristics
    • Supported Jetpack Versions
  • TX1
    • Characteristics
    • Supported Jetpack Versions
• Contact Us

Libraries

NVIDIA Jetpack includes several libraries for applications in deep learning, GPU accelerated applications, and computer vision. These libraries includes:

• TensorRT: A high-performance deep learning inference runtime for image classification, segmentation, and object detection neural networks. It speeds up deep learning inference as well as reducing the runtime memory footprint for convolutional and deconv neural networks.

• CUDA-X: A set of libraries that deliver higher performance, compared to CPU-only alternatives acroos application domains, including AI and high-performance computing. Some of these libraries includes:
  • CUDA Math Libraries
  • Parallel Algorithm Libraries
  • Data Processing Libraries
  • Communication Libraries
  • Quantum Libraries
  • Partner Libraries
  • Computational Lithography
  • Image and Video Libraries
  • Deep Learning Core

• CUDA Deep Neural Network (CuDNN): Is a library that provides high-performance primitives for deep learning frameworks. It includes support for convolutions, activation functions, and tensor transformations.

• Multimedia API: The Jetson Multimedia API package provides low level APIs for flexible application development.
  • Camera application API: libargus offers a low-level frame-synchronous API for camera applications, with per frame camera parameter control, multiple (including synchronized) camera support, and EGL stream outputs.
  • Sensor driver API: V4L2 framework enables video decode, encode, format conversion and scaling functionality. V4L2 for encode opens up many features like bit rate control, quality presets, low latency encode, temporal tradeoff, motion vector maps, and more.

• Vision Programming Interface (VPI): Is a software library that provides Computer Vision / Image Processing algorithms implemented on PVA (Programmable Vision Accelerator), GPU and CPU.

• OpenCV: Open source library for computer vision, image processing and machine learning, and now features GPU acceleration for real-time operation.

< >