In this modern AI universe, each and every industry is trying to build an Artificial Intelligence in their fields. For AI research, lot of industries introducing AI development boards.
Among those industry, well known industry who is popular on GPU’s (Graphical Processing Unit) introduced low cost GPU board suitable for AI applications, which includes Machine learning and deep learning.
- Nvidia introduced the low cost GPU boards which is capable to perform heavy AI applications, in tiny board with very low power consumption.It is helpful for researchers, hobbyist to develop AI based applications.
- So AI begins with Machine learning and deep learning sort of applications can be easily done by using this Nvidia Nano.
- It also has SD Card image boot like Raspberry pi, but with GPU. So that heavy applications can be performed using this Nvidia Jetson Nano
- Though its features are high, it requires less power. Which can be powered up by simply using 5 volt with 2 Amps adapter.
- Jetson Nano is used for various ML & DL research in various fields for Image processing, Speech processing and sensor based analysis.
- 128-core Maxwell GPU
- Quad-core ARM A57 @ 1.43 GHz CPU
- 4 GB 64-bit LPDDR4 25.6 GB/s Memory
- Storage is based on SD Card used (32 GB high Class SDCard Recommended)
- 4K @ 30, 4x 1080p @ 30, 9x 720p @ 30 (H.264/H.265) Video Encode
- 1x MIPI CSI-2 DPHY lanes Camera
- 4K @ 60 | 2x 4K @ 30 | 8x 1080p @ 30 | 18x 720p @ 30 (H.264/H.265) Video Decode
- Gigabit Ethernet
- HDMI 2.0 and eDP 1.4 Display
- 4x USB 3.0 USB
- USB 2.0 Micro-B Supply USB
- 40 Pins GPIO with UART, SPI, I2C and I2S
Easy to use Nano
If you are familiar in using Raspberry Pi, it is almost same on using Nvidia Nano.
- Get the Nvidia Jetson Nano with 32 GB class 10 SDCard
- Download the boot image from here
- Boot the image into the SD card. To know about booting image into SDCard Clickhere
- It is recommended to use 5 Volt with 2 Amps USB Charger or else it can be powered up by using Adapter or GPIO
- Connect HDMI monitor with the Jetson Nano, no other converter is recommended like VGA to HDMI (Only HDMI input Monitor)
- Jetson nano doesn’t has inbuilt WiFi like Raspberry Pi, so you can provide the Network by connecting the Ethernet or by giving USB Modem.
- Choose the language, country, clock settings and every other setup in first boot
- Now its ready for your AI program.
Programming and Installing Libraries in Jetson Nano
Programming Nano is simple as same as Raspberry Pi,
- Nvidia Nano doesn’t has IDLE with first boot, but we can install IDLE for python3 by entering below command in Terminal
sudo apt-get install idle3
- So before programming, libraries are needed to be installed. Since image provided for Jetson is Ubuntu , we can install libraries using pip commands and installing applications using apt-get command.
- Python programs i.e, .py files can be run through IDLE by pressing F5 or we can run through terminal using command. Ex. sudo python3 led.py for Python3 and sudo python led.py for Python2
- It also has 40 GPIO pins with same order in Raspberry Pi 3. We can use for hardware applications with image related applications.
- Nano has CSI Camera port, you can use CSI Camera made for Jetson as well as you can use Raspberry Pi CSI Camera.
- Since it has USB ports you can use USB Camera and also by connecting nano with the Network you can use your Android Mobile phone for Image processing application, when you need good quality of image processing but didn’t have camera.