A 10.7-µJ/Frame 88% Accuracy CIFAR-10 Single-Chip Neuromorphic Field-Programmable Gate Array Processor Featuring Various Nonlinear Functions of Dendrites in the Human Cerebrum

A neuromorphic architecture is suitable for low-power tiny-machine learning processors. However, the large number of synapses utilized in recent deep neural networks require multichip implementation, resulting in large power consumption due to chip-to-chip interfaces. Here, we present a 10.7-µJ/frame single-chip neuromorphic field-programmable gate array (FPGA) processor. To reduce the required hardware (HW) resources, we have developed two techniques. The first is a dendrite-inspired nonlinear neural network that mimics various nonlinear functions of dendrite spines in the human cerebrum. The second is a line-scan-based architecture that reduces the total amount of HW resources. The 14-layer convolutional neural network (CNN), which achieves an 88% accuracy with the CIFAR-10 dataset, was implemented on a single FPGA board. Compared to a state-of-the-art spiking CNN-based neuromorphic FPGA processor, the energy efficiency of the proposed architecture is improved by a factor of 94.4 while achieving a 6% better classification accuracy.