Neural Computing Development Kit

Neural Computing Development Kit

Regular price
Sold out
Sale price
$149.99

The Atomic Pi is perfect for supporting the Movidius inference engine.  Included in the kit is everything you need to explore AI, Neural Computing, the Movidius VPU and Machine Vision. Learn tomorrow's technology today.

Some use cases demand custom I/O and low power consumption; others need storage and processing capabilities. With this kit, you get both on a flexible Linux system. The NCS2 processes video using energy efficient deep learning, freeing up the CPU for other tasks including your own custom apps.  Integration into generic monitoring frameworks via gstreamer, MQTT and other protocols is easy.

Open the box, pop in the USB stick, load the machine vision demos and explore Neural Computing Technology on your workstation.  Customize. Copy your models over to the APi and run them out of the box on a reliable, power-efficient Atom SBC with a ready-to-use camera attached.

  • A brief but detailed quick-start / explanation guide to get you up and running.
  • Atomic Pi SBC with 16GB Emmc.
  • Small breakout board and 5V 5A power supply. 
  • Matching machine vision camera and cables. 
  • Neural Compute Stick 2 - Think of it as a hardware accelerator.
  • Ubuntu and OpenVino™ come pre-loaded on the A Pi with patches where necessary.
  •  Batteries are included: Code samples come with select pre-downloaded and pre-optimized models to get you started in no time.
  • All the hardware and software tweaks to make it a reliable SBC AI platform.
  • We currently plan to ship with OpenVINO and gstreamer pre-built as well as a number of samples loaded on the A Pi flash.
  • Once you have the kit you can choose to install the full Intel® Distribution of OpenVINO™ toolkit.

 

 

Additional parts you may need:

  • A separate PC for experimenting with models.
  • An enclosure for your embedded system.
  • A USB hub, keyboard and mouse to set up and configure the A Pi.
  • Add an antenna if you'd like long range WiFi.