Computation Error Analysis of Block Floating Point Arithmetic Oriented Convolution Neural Network Accelerator Design

Open in new window