The world’s leading publication for data science, AI, and ML professionals.

Jetson Nano 2GB for AI & IoT projects

A brief introduction of NVIDIA Jetson Nano 2GB Developer Kit

Photo from the Author.
Photo from the Author.

On 5 October 2020, NVIDIA launched Jetson Nano 2G Developer Kit, a small, powerful 59 American Dollars computer for learning and developing AI, IoT, and robotics applications. In this post, I will guide you on how to install and configure it.

If you don’t know about Jetson Nano 2GB, I recommend you to see the Getting Started with Jetson Nano 2GB Developer Kit video on the NVIDIA Developer YouTube channel:


The Setup

You will need access to a personal computer and other accessories not included with the $59 Jetson Nano 2GB Developer Kit.

For the basic setup, I am using the following configuration:

Considerations

There are some general recommendations detailed below.

Photo by the Author.
Photo by the Author.

The Power Supply

The Jetson Nano 2GB Developer Kit requests a USB-C of 5V⎓3A power supply; you can check the supported power supplies.

It does not support the USB-C Power Delivery protocol, meaning that it will not alert you or fail, but when the power is not enough will shut down or something else.

Initially, I used a standard USB-C from Amazon Charger 5.2V⎓1.8A (9W), which works. Still, I followed the NVIDIA recommendation and changed to CanaKit with a proper 5V⎓3A power supply for future projects.

The Network

The Jetson Nano, 2GB Developer Kit, doesn’t come with a Wireless or Bluetooth integration; instead, you can use a USB dongle connecting to the USB 3.0 port for better performance. It suggests using an extension cable to reduce EMI interference between the USB networking adapter and the developer kit.

Not all regions include an external USB 802.11ac wireless adapter and extension cable in the Developer Kit.

I am using USB dongle Edimax 2-in-1 Wi-Fi 4 802.11n N150 + Bluetooth Low Energy (BLE) 4.0

The Temperature

Parallel processing may become hot; never touch the heatsink (big black metal) during or after use.

It is a good idea to check the temperature and power supply with the TegraStats utility command:

sudo tegrastats

Step 1: Build a Linux MicroSD Card Image

Initially, used an external computer to begin the setup, I am using Windows 10, but the steps are similar if you are using Mac or Linux.

  1. Download the Jetson Nano 2GB Image
  2. Connect your microSD to your computer
  3. Do a quick format with a blank label on your microSD card with a format utility.
Screenshot from the Author.
Screenshot from the Author.
  1. Install Balena Etcher, choose from the downloaded image file, and flash it. This process can take a while.
Screenshot by the Author.
Screenshot by the Author.

Step 2: Configuring Linux

Insert the microSD into the Jetson Nano with the metal up to the heatsink

Photo by the Author
Photo by the Author

Connect the HDMI Monitor and USB Wireless to the USB 3.0 port, USB Keyboard, USB mouse, and at the end, the USB-C power supply.

If you don’t have Mouse/Keyboard and HDMI Monitor, you can try the Headless setup with a USB/MicroUSB cable.

Photo by Author
Photo by Author

After a minute, you will see a Linux Lightweight Ubuntu 18.06

Photo by the Author
Photo by the Author

In my case, It did not recognize my USB Razer mouse, and I had to continue using only the keyboard by pressing the Tab key.

After boot, the first that you will see is a request to accept the EULA (End User License Agreement)

Accept EULA. Photo by the Author
Accept EULA. Photo by the Author

The next step is to choose the language for the Ubuntu Linux System configuration.

Configure Language. Photo by the Author
Configure Language. Photo by the Author

Choose your Keyboard language.

Configure Keyboard. Photo by the Author
Configure Keyboard. Photo by the Author

Configure your WI-FI connection

Configure Wi-FI. Photo by the Author
Configure Wi-FI. Photo by the Author

Set the Timezone.

Configure Timezone. Photo by the Author
Configure Timezone. Photo by the Author

Create your administrator user

Creating an admin user. Photo by Author
Creating an admin user. Photo by Author

Configure the partition size; the recommendation is to use the default value.

Adding partition size. Photo by the Author
Adding partition size. Photo by the Author

Create a SWAP file partition of at least 4GB

Create a Swap file. Photo by the Author
Create a Swap file. Photo by the Author

Step 3: Enabling Remote Access

Virtual Network Computing (VNC)

VNC Server is installed but, by default, is turned off and not configured. Run these commands to configure and restart the device

sudo apt-get update
mkdir -p ~/.config/autostart
cp /usr/share/applications/vino-server.desktop ~/.config/autostart/.
gsettings set org.gnome.Vino prompt-enabled false
gsettings set org.gnome.Vino require-encryption false
#Replace thepassword with your desired password
gsettings set org.gnome.Vino authentication-methods "['vnc']"
gsettings set org.gnome.Vino vnc-password $(echo -n 'thepassword'|base64)
sudo reboot

Get the IP Address to be used by the VNC Viewer from your computer. You can get the Viewer here.

ifconfig | grep inet
WSL Ubuntu Prompt. Screenshot by Author.
WSL Ubuntu Prompt. Screenshot by Author.
Real VNC Viewer. Screenshot by the Author.
Real VNC Viewer. Screenshot by the Author.

Secure Shell (SSH)

SSH Server, by default, is active in the Jetson Nano. To connect, you can use PowerShell prompt, Windows Subsystem Linux (WSL), or PuTTY from your computer with the following command:

ssh <user>@<ip address>
Power Shell Prompt. Screenshot by the Author.
Power Shell Prompt. Screenshot by the Author.

If you don’t have WSL2 and want to use it, I recommend checking my previous post, Configuring Jupyter Notebook in WSL2.

WSL Ubuntu Prompt. Screenshot by Author.
WSL Ubuntu Prompt. Screenshot by Author.

Conclusions

NVIDIA launched an affordable $59 device that, I think, will change how we learn and develop Artificial Intelligence (AI) and the Internet of Things (IoT) projects because of the combination of Ubuntu Linux and the power of NVIDIA hardware in a small device.


What is next?

If you are interested in this topic, I recommend you check the Jetson AI Certification program and the Education projects.


Helpful resources

Jetson Nano 2GB Developer Kit

Jetson Nano 2GB Developer Kit User Guide

Introducing the Ultimate Starter AI Computer, the NVIDIA Jetson Nano 2GB Developer Kit | NVIDIA…

Jetson Download Center

New Jetson Nano 2GB Developer Kit

Two Days to a Demo

Jetson AI Courses and Certification

Jetbot

dusty-nv/jetson-inference


Related Articles