An Ultra-Low-Power Serial Implementation for Sigmoid and Tanh Using CORDIC Algorithm

Abstract

Activation functions (AFs) such as sigmoid and tanh play an important role in neural networks (NNs). Their efficient implementation is critical for always-on edge devices. In this work, we propose a serial-arithmetic architecture for AFs in edge audio applications using the CORDIC algorithm. The design enables to dynamically trade-off throughput/latency and accuracy, and pos-sesses higher area and power efficiency compared to conventional methods such as look-up table (LUT) and piece-wise linear (PWL)-based methods. Considering the throughput difference among the designs, we evaluate average power consumption taking into account active and idle working cycles for same applications. Synthesis results in a 22nm process show that our CORDIC-based design has an area of 545.77 μm2 and an average power of 0.69 μW for a keyword spotting task, achieving a reduction of 36.92% and 71.72% in average power consumption compared to LUT and PWL-based implementations, respectively.