Development board ArduinoIn its various versions, it is a great option for makers and developers who want to create their own DIY projects. On the other hand you also have the Raspberry Pi, a small and cheap SBC to create a multitude of projects. To all this is added the enormous number of alternatives on the market. But you are probably looking for something more specific, something to work with neural networks and AI. Then the NVIDIA Jetson Nano board is what you are looking for.
If you plan to make developments based on neural networks or learn about it, you can count on this board Nvidia Jetson Nano for it. All for just over € 100, not too high considering prices of other smart systems ...
What is Jetson?
Nvidia Jetson Nano is a project of the well-known graphics chip company that enables the development of many new small-size AI systems. In addition, looking at the prices of other projects for neural networks, it does so at a price and a fairly low energy consumption.
With this development board you can create multitude of projects, such as IoT applications, from small domestic robots, to other intelligent systems that you can think of, through network video recorders (NVRs), intelligent gateways, etc.
All in a small PCB module of about 70x45mm in size in its Nano version, the smallest. In fact, it is a SOM type module (System on Module) ready to work with it, together with the development kit.
Technical characteristics
As for the the technical characteristics From the NVIDIA Jetson Nano, you are going to find yourself with a powerful board capable of developing 472 GFLOPs of performance to execute modern AI algorithms very quickly. It could even run multiple neural networks in parallel and process multiple high-resolution sensors simultaneously.
And all that with a fairly efficient energy consumption. Despite its power, only consumes between 5 and 10w. A fairly low power considering the benefits it offers.
For more details, I leave you the complete table of specs:
| GPU | NVIDIA Maxwell architecture™ with 128 NVIDIA CUDA cores® | |
| CPU | ARM processor® Cortex®-A57 MPCore Quad Core | |
| Conference proceedings | 4GB 4-bit LPDDR64 | |
| Storage | 16GB eMMC 5.1 Flash storage | |
| Video encoding | 4K 30 frames (H.264 / H.265) | |
| Video decoding | 4K 60 frames (H.264 / H.265) | |
| Camera | 12-way (3 x 4 or 4 x 2) MIPI CSI-2 DPHY 1.1 (18 Gbps) | |
| Connectivity | Gigabit Ethernet | |
| Screen | HDMI 2.0 or DP 1.2 | eDP 1.4 | DSI (1 x 2) 2 simultaneous | |
| UPHY | 1 1/2/4 PCIE, 1 USB 3.0, 3 USB 2.0 | |
| IS | 1 SDIO / 2 SPI / 4 I2C / 2 I2S / GPIO | |
| Size | 69,6 mm x 45 mm | |
| Mechanics | 260 pin connector | |
Other similar products
NVIDIA also offers other similar products to Jetson Nano for AI development with artificial neural networks. Some examples are:
- Jetson Xavier NX: a SOM module that offers all the power of a supercomputer with very small dimensions. You can get up to 21 TOPs, that is, 21 Tera Operations per seconds. More than enough power to run modern neural networks in parallel and process data from multiple high-resolution sensors at the same time.
- Jetson AGX Xavier: a new module that marks a milestone in terms of computational density and efficiency. For AI, allowing the creation of new generations of intelligent machines.
- Jetson TX2- Another high-speed, energy-efficient development board for embedded AI computing. A supercomputer in a module based on the NVIDIA Pascal GPU. With up to 8GB of RAM and a bandwidth of 59,7GB / s.
Of course, as you can see, his older brothers have Prices considerably higher ...
Acquire the NVIDIA Jetson Nano
If you decide to buy the NVIDIA Jetson Nano, you have several possibilities for it. One is the products offered through the Amarzon platform. You will find both the development board without more, or more complete development kits that include the power adapter, etc. For example:
- NVIDIA Jetson Nano Basic
- NVIDIA Jetson Nano Kit with power adapter, 64GB microSD, USB
- Buy only the SOM module
Remember that the technologies of machine learning algorithm , artificial intelligence, deep learning, etc., are increasingly "fashionable", since they have a multitude of very interesting applications. Therefore, learning about them could be interesting in order to create new projects for the future or get interesting jobs in companies that require this type of skills ...