A Sliding‐Kernel Computation‐In‐Memory Architecture for Convolutional Neural Network

Presently described is a sliding‐kernel computation‐in‐memory (SKCIM) architecture conceptually involving two overlapping layers of functional arrays, one containing memory elements and artificial synapses for neuromorphic computation, the other is used for storing and sliding convolutional kernel matrices. A low‐temperature metal‐oxide thin‐film transistor (TFT) technology capable of monolithically integrating single‐gate TFTs, dual‐gate TFTs, and memory capacitors is deployed for the construction of a physical SKCIM system. Exhibiting an 88% reduction in memory access operations compared to state‐of‐the‐art systems, a 32 × 32 SKCIM system is applied to execute common convolution tasks. A more involved demonstration is the application of a 5‐layer, SKCIM‐based convolutional neural network to the classification of the modified national institute of standards and technology (MNIST) dataset of handwritten numerals, achieving an accuracy rate of over 95%. The sharing of convolution kernels in convolutional operations creates a significant asymmetry between data and weight (kernel) volume. This study optimizes convolutional operations at the hardware level by utilizing threshold adjustments in dual‐gate thin‐film transistors (TFTs). Achieved is a reduction in memory access operations of up to 88% through data preloading and the sliding of small kernels.