Intel has released a new video which outlines some of the key details for its Movidius Neural Compute Stick.
The Movidius Neural Compute Stick from Intel is basically a tiny fanless deep learning device that can be used for AI programming at the edge. Powered by Intel Movidius Vision Processing Unit (VPU), the same one that is used for smart security cameras, gesture-controlled drones, industrial machine vision equipment, and more, it enables AI applications that are not reliant on a connection to the cloud.
The Movidius Neural Compute Stick (NCS) measures at 72.5 x 27 x 14mm and uses USB 3.0 Type-A connection. It requires an x86_64 computer running Ubuntu 16.04 or Raspberry Pi 3 Model B running a Stretch desktop or Ubuntu 16.04 VirtualBox instance, at least 1GB of RAM and 4GB of storage space.
The Movidius NCS, in combination with Intel Movidius Neural Compute SDK, allows developers to deploy Convolutional Neural Network (CNN) on low-power applications.
You can check out more about the Intel Movidius Neural Compute Stick over at Intel’s dedicated Movidius developer site.