An Energy-Efficient Adiabatic Capacitive Neural Network Chip

TL;DR

This paper introduces an energy-efficient mixed-signal adiabatic neural network chip that achieves high classification accuracy and significant energy savings suitable for edge devices.

Contribution

The paper presents a novel 130nm CMOS chip with adiabatic logic and multiple multiply-accumulate units for low-power image classification.

Findings

Achieves over 95% accuracy on 8x8 1-bit image classification

Realizes 2.1x to 6.8x energy savings over traditional CMOS implementations

Operates reliably within 2.7% of software-based classification performance

Abstract

Recent advances in artificial intelligence, coupled with increasing data bandwidth requirements, in applications such as video processing and high-resolution sensing, have created a growing demand for high computational performance under stringent energy constraints, especially for battery-powered and edge devices. To address this, we present a mixed-signal adiabatic capacitive neural network chip, designed in a 130$nm$ CMOS technology, to demonstrate significant energy savings coupled with high image classification accuracy. Our dual-layer hardware chip, incorporating 16 single-cycle multiply-accumulate engines, can reliably distinguish between 4 classes of 8x8 1-bit images, with classification results over 95\%, within 2.7\% of an equivalent software version. Energy measurements reveal average energy savings between 2.1x and 6.8x, compared to an equivalent CMOS capacitive implementation.