A 183.4nJ/inference 152.8μW Single-Chip Fully Synthesizable Wired-Logic DNN Processor for Always-On 35 Voice Commands Recognition Application

A 183.4nJ/inference single-chip wired-logic DNN processor that is capable of recognizing all 35 commands defined in the industrial standard voice recognition data set (Google Speech Command Dataset) is developed. The fully synthesizable processor recognizes 3.5 times more commands with 1.6 times better energy efficiency than the state-of-the-art analog processor while keeping design cost low. By implementing all the processing circuits and wiring required for the 16-layer DNN onto a single chip (7.63 mm
²
in 40nm), the need to store weight coefficients and intermediate data in DRAM/SRAM is eliminated. Owing to the proposed architecture, a low power consumption of 152.8$\mu$W is achieved, which is low enough for always-on applications on battery-powered IoT devices.