Zebra offers users FPGA-based class-leading acceleration for neural network inference. The user-defined neural network works on Zebra just as it would on GPU or CPU. Zebra is fully integrated with the traditional Deep Learning infrastructures, like Caffe, MXNet or TensorFlow.

Zebra includes the FPGA image and software stack. There is zero FPGA knowledge required nor a single line of code to write to use Zebra. Simply link to the Zebra library to switch from CPU or GPU to FPGA in minutes and accelerate your neural network inference.

*** This AMI is to be used for non-commercial use ***

This AMI contains the following components:
Ready-to-run demonstration of image classification using CaffeNet.This AMI includes a pre-trained ready-to-use CaffeNet neural network.
This AMI allows you to run inference on your images.
Supports 1 FPGA (compatible with f1.2xlarge).
This AMI does not support any neural network infrastructure. Contact us for more information on how to link to Zebra.
Includes Zebra FPGA and software stack - No need to compile an FPGA or to know there is one.

For more details :
/home/centos/Mipsology/doc/README.md
http://www.mipsology.com/aws/getting_started.html

If you are looking for commercial Zebra, look for other Zebra AMI on AWS Marketplace.