Movidius™ Neural Compute SDK

Home of the ncsdk used to develop software for the Movidius™ Neural Compute Stick

View the Project on GitHub


The Movidius™ Neural Compute SDK and Movidius™ Neural Compute Stick (NCS) enables rapid prototyping, validation and deployment of Deep Neural Networks (DNNs.)

The NCS is used in two primary scenarios:

The following diagram shows the typical workflow for development with the NCS:

The training phase does not utilize the NCS hardware or SDK, while the subsequent phases of “profiling, tuning and compiling” and “prototyping” do require the NCS hardware and the accompanying Movidius™ Neural Compute SDK

The SDK contains a set of software tools to compile, profile, and check validity of your DNN as well as an API for both the C and Python programming languages. The API is provided to allow users to create software which offloads the neural network computation onto the Movidius™ Neural Compute Stick.

Here is more information on the architecture of the Neural Compute Stick.


Neural Compute SDK currently supports two Deep Learning frameworks.

  1. Caffe : Caffe is a deep learning framework from Berkeley Vision Labs.
  2. TensorFlow™: TensorFlow™ is a deep learning framework from Google.

See how to use networks from these supported framework with NCS.

Installation and Examples

The following commands install NCSDK and run examples. Detailed instructions for installation and Configuration

git clone && cd ncsdk && make install && make examples

Movidius™ Neural Compute SDK Tools

The SDK comes with a set of tools to assist in development and deployment of applications that utilize hardware accelerated Deep Neural Networks via the Movidius™ Neural Compute Stick (NCS). Each tool and its usage is described on this page below

Neural Compute API

Applications for inferencing with Neural Compute SDK can be developed either in C/C++ or Python. The API provides a software interface to Open/Close Neural Compute Sticks, load graphs into the NCS and run inferences on the stick.

Movidius™ Neural Compute User Forum

There is an active user forum in which users of the Neural Compute Stick discuss ideas and issues they have with regard to the NCS. The link to the NCS User Forum is:

The forum is a good place to go if you need help troubleshooting an issue. You may find other people that have figured out the issue or get ideas for how to fix it. The forum is also monitored by Movidius™ engineers which provide solutions as well.


There are several examples including the following at the github

The examples demonstrate compiling, profiling and running inferences using the network on the Movidius™ Neural Compute Stick. Each example contains a Makefile. Running ‘make help’ in the example’s base directory will give possible make targets.

git clone # Already done during installation
(cd ncsdk/examples && make) # run all examples
(cd ncsdk/examples/caffe/GoogLeNet && make) # Run just one example

Release Notes