zero value Method and apparatus for processing computation of zero value in processing of layers in a neural network

摘要

A method and apparatus for processing layers of a neural network include fetching IFM tiles of an input feature map (IFM) tensor and kernel tiles of a kernel tensor, and IFM tiles using IFM sparsity and kernel sparsity. And a convolution operation on the kernel tiles, and generates a plurality of OFM tiles corresponding to the IFM tiles.