Getting Started with Nvidia Jetson Nano and installing Tensorflow GPU

As a deployment engineer at Nsemble.ai , i work on researching and delivering Cloud and IOT based machine learning and deep learning solutions. Deploying an object detection model on an IOT device is a tedious task. Even the high end deep learning model process slow on basic devices like Raspberry pi. I researched around for a while and got this amazing device fit for production ready called Nvidia Jetson Nano. Its graphics experience is amazing and it is also cost effective ($99). This iot device is perfect for deploying the heavy deep learning models.

This is a simple blog for getting started with Nvidia Jetson Nano IOT Device (Device Overview and OS Installation) followed by installation of the GPU version of tensorflow.

Introduction

The NVIDIA® Jetson Nano™ Developer Kit is a small AI computer for makers, learners, and developers. After following along with this brief guide, you’ll be ready to start building practical AI applications, cool AI robots, and more.

Nvidia Jeston Nano newly opened

Prepare for Setup

Items for Getting Started

microSD Card

The Jetson Nano Developer Kit uses a microSD card as a boot device and for main storage. It’s important to have a card that’s fast and large enough for your projects; the minimum recommended is a 16GB UHS-1 card.

See the instructions below to flash your microSD card with operating system and software.

Micro-USB Power Supply

You’ll need to power the developer kit with a good quality power supply that can deliver 5V⎓2A at the developer kit’s Micro-USB port. Not every power supply rated at “5V⎓2A” will actually do this.

Optional Items

Wireless Networking Adapter

Jetson Nano Developer Kit includes a gigabit Ethernet port, but also supports many common USB wireless networking adapters, e.g., Edimax EW-7811Un.

Write Image to the microSD Card

To prepare your microSD card, you’ll need a computer with Internet connection and the ability to read and write SD cards, either via a built-in SD card slot or adapter.

1. Download the Jetson Nano Developer Kit SD Card Image, and note where it was saved on the computer.
2. Write the image to your microSD card by following the instructions below according to the type of computer you are using Windows.

Format your microSD card using SD Memory Card Formatter from the SD Association.

SD Formatter

1. Download, install, and launch SD Memory Card Formatter for Windows.
2. Select card drive
3. Select “Quick format”
4. Leave “Volume label” blank
5. Click “Format” to start formatting, and “Yes” on the warning dialog

Use Etcher to write the Jetson Nano Developer Kit SD Card Image to your microSD card

1. Download, install, and launch Etcher.

Etcher overview

Etcher is used to make the sd card bootable using the os provided for ready to go ubuntu desktop provided by nvidia developers for jetson nano

1. Click “Select image” and choose the zipped image file downloaded earlier.
2. Insert your microSD card if not already inserted.

Click cancel if windows prompts you with a dialog like this:

4. Click “Select drive” and choose the correct device.

5. Click “Flash!” It will take Etcher about 10 minutes to write and validate the image if your microSD card is connected via USB3.

6. After Etcher finishes, Windows may let you know it doesn’t know how to read the SD Card. Just click Cancel and remove the microSD card.

Click cancel

Setup and First Boot

Setup Steps

1. Unfold the paper stand and place inside the developer kit box.
2. Set the developer kit on top of the paper stand.
3. Insert the microSD card (with system image already written to it) into the slot on the underside of the Jetson Nano module

4. Power on your computer display and connect it.

5. Connect the USB keyboard and mouse.

6. Connect your Micro-USB power supply (5V⎓2A). The Jetson Nano Developer Kit will power on and boot automatically. The first output on monitor will be like this as soon as nvidia boots.

First Boot

A green LED next to the Micro-USB connector will light as soon as the developer kit powers on. When you boot the first time, the Jetson Nano Developer Kit will take you through some initial setup, including:

• Review and accept NVIDIA Jetson software EULA
• Select system language, keyboard layout, and time zone
• Create username, password, and computer name
• Log in

After Logging In

You will see this screen. Congratulations!

Ubuntu Desktop as soon as system boots up on Nvidia Jetson Nano

Troubleshooting

Power

If you cannot boot your Jetson Nano Developer Kit, the problem may be with your USB power supply.Please use a good quality power supply like this one.

It’s also important to have a good quality cord connecting your power supply to the developer kit:

• It’s good to use a power supply with permanently attached cord.
• Shorter cables will drop less voltage.

Display

HDMI to DVI adaptors are not supported. Please use a display that accepts HDMI or DP input.

Installing Tensorflow GPU on Nvidia Jetson Nano

The reason why i have included this step is because its pretty easy to install the GPU version on this device.

Open terminal and enter following commands:

Install the dependencies:

sudo apt-get install python3-pip libhdf5-serial-dev hdf5-tools

Install tensorflow-gpu library

sudo pip3 install --pre --extra-index-url https://developer.download.nvidia.com/compute/redist/jp/v42 tensorflow-gpu

In case you get HTTP 404 Not found error then try the following command

pip3 install — pre — extra-index-url https://developer.download.nvidia.com/compute/redist/jp/v42 tensorflow-gpu

If you enjoy my blog and would like to support my work, consider buying me a coffee. Your contribution will help me keep the blog running and creating valuable content. Thank you!