CONFIGURABLE PROCESSOR ELEMENT ARRAYS FOR IMPLEMENTING CONVOLUTIONAL NEURAL NETWORKS

摘要

An exemplary apparatus disclosed herein includes an array of processor elements, the array including rows each having a first number of processor elements and columns each having a second number of processor elements. The disclosed exemplary apparatus also includes configuration registers for storing descriptors for configuring an array to implement a layer of a convolutional neural network based on a data flow schedule corresponding to a tensor processing template of a plurality of tensor processing templates, , the processor elements of the processor elements operate on input activation data and filter data associated with the layer of the convolutional neural network to generate output activation data associated with the layer of the convolutional neural network, based on the descriptors. is configured to implement the tensor processing template of The disclosed exemplary apparatus further includes a memory for storing input activation data, filter data, and output activation data associated with the layers of the convolutional neural network.