📖
Wiki
Back to my personal website
  • Home
  • Equipment and Devices
    • 3D Printer
    • Laser Cutter
    • Motion Capture System
    • Sensors
      • RGB-D Cameras
      • Velodyne LiDAR
      • Zed Camera
      • RealSense D435i
      • IMU
    • eGPU
    • Nvidia AGX Xavier
    • CPU Benchmark
    • Installation Checklist
  • Development
    • Linux
      • Shell
      • GDB
      • Git
      • Tmux
      • Network
      • Tricks
      • Debug FAQ
    • CMake
      • Catkin Tools
      • CMakeLists
      • CMake Variables
      • CMake Commands
      • CMake: find_package()
    • ROS
      • Gazebo
      • wstool
      • roslaunch
      • rosbag
      • multi-threaded spinner
    • ROS2
      • Convert ROS1 bag to ROS2 bag
    • C++
      • C++ 11
      • C++ Examples
      • C++ Debug
      • Factory Method
      • Timing
    • Google Tools
      • GLog
      • GFlags
      • GTest
      • Style Guide
      • Clang Format
    • PCL
      • Point Type
      • Methods
      • Architecture
      • Code Explained
    • Open3D
      • Python API
      • Registration
      • Visualization
      • Tools
    • OpenCV
      • Documentation
      • Modules
    • Other Libraries
      • Eigen
      • Ceres
      • g2o
      • GTSAM
    • Website
  • Algorithm
    • SLAM
      • K-D Tree
      • Octree
      • Bag of Words
      • Distance Measures
      • Coordinate Systems
      • LOAM
      • Iterative Closest Point
      • Generalized ICP
      • Mahalanobis Distance
    • Computer Science
      • Computational Model
      • Sorting
      • Analysis
      • Complexity Classes (P, NP)
      • Divide and Conquer
      • Greedy Algorithm
      • Dynamic Programming
      • Tree
      • Graph
    • Computer Vision
      • Camera Models
      • Distortion
      • Motion Models
      • Shutter
      • Image Sensors
      • Epipolar Geometry
      • Multiple-View Geometry
    • Datasets
      • RGB-D Datasets
      • Point Cloud Datasets
      • LiDAR SLAM Datasets
  • Math
    • Optimization
      • Convex Optimization
      • Descent Methods
    • Probability
      • Moment
      • Covariance Matrix
      • Stochastic Process
    • Topology
      • References
      • Concepts
      • Topological Spaces
      • Representation of Rotations
      • Representation of 3-sphere
    • Algebra
      • Linear Algebra
      • Matrix Factorization
      • Condition Number
      • Matrix Lie Group
    • Differential Geometry
      • Manifold
      • Submanifold
      • Quotient Manifolds
      • Tangent Space
  • Quadrotor
    • PX4 Development
    • Companion Computer
    • Drone Hardware
    • Propeller Lock
    • Debug
  • Favorites
    • Bookmarks
Powered by GitBook
On this page
  • Introduction
  • Installation
  • Post Installation

Was this helpful?

  1. Equipment and Devices

Nvidia AGX Xavier

PreviouseGPUNextCPU Benchmark

Last updated 1 year ago

Was this helpful?

Introduction

was released in 2018 and is the most powerful onboard computer so far. In early 2020, Nvidia has doubled its RAM (from 16G to 32G) at the same price. Some notable features are as follows.

  • 8-core ARMv8 CPU

  • 512 CUDA cores

  • 32GB DDR RAM

  • 32GB eMMC storage

  • possible to add 1TB NVMe SSD

  • acceptable weight (700g)

  • price $699

See and for more information regarding interfaces and features. See for more details.

Purchase List:

Installation

It is required to have another Ubuntu 16 or 18 system to help flash OS image into AGX Xavier. It is recommended to use natively installed Ubuntu, but I have successfully flashed AGX Xavier using a Ubuntu 18 VM in VMware.

Steps to install the system:

  • Press force recovery button and power button on AGX Xavier to enter recovery mode. To verify, run lsusb command on host Ubuntu and see if a Nvidia device is available. (Only in recovery mode can Xavier be detected as a device via USB connection.)

  • Follow the instructions on SDK Manager and finish installation. Note that SDK Manager tries to install everything at once. However, I recommend to decompose it into three steps.

    • flash Jetson OS first (uncheck SDK components; network connection on Xavier is not required at this step)

    • once finished, unplug USB connection, restart Xavier in normal mode, connect mouse, keyboard and monitor and set up this new Ubuntu system (username, password, time zone, etc.)

    • connect Xavier and host Ubuntu to the same network, restart SDK Manager on host Ubuntu and install Jetson SDK Components only (uncheck Jetson OS)

Post Installation

sudo nvpmodel -q     # check status
sudo nvpmodel -m 0   # switch to max performance mode

Next, we can proceed to install SSD storage and move our entire OS from eMMC to the new SSD storage.

// make sure SSD is formatted to ext4, and a partition (p1) is set
git clone https://github.com/jetsonhacks/rootOnNVMe.git
cd rootOnNVMe
./copy-rootfs-ssd.sh
./setup-service.sh
sudo reboot
sudo apt install linux-firmware

In host Ubuntu system (16 or 18), download and install . You may need to register an account in order to have access to the download page.

Follow or to set up cable connections. (refer to hardware connections only; the installer in setup video is no longer available; we use sdkmanager instead.)

As introduced in the , Nvidia AGX Xavier by default runs at 15W balance mode with 4 core CPU. To check status and switch model, run the following code.

To install NVMe SSD, see for hardware instructions. Software instructions may differ in the last a couple steps. I recommend leaving 32GB unallocated for swap.

See and this for instructions regarding moving root filesystem from eMMC to SSD storage. The original discussion on Nvidia forum can be seen . In summary, the commands are as follows.

Lastly, you may want to install the WiFi + Bluetooth module. Hardware installation is obvious I believe. Software side only the following command is required, though they should be already installed. See also .

Nvidia Jetson AGX Xavier
initial setup video
unboxing video
user manual
Nvidia AGX Xavier (32G)
1TB Samsung 970 EVO SSD
8265AC WiFi and Bluetooth module with antenna
Nvidia SDK Manager
initial setup video
user manual
unboxing video
this video
this page
github repo
here
this page