Realization and training of an inverter-based printed neuromorphic computing system

Dennis D Weller,Michael Beigl,Michael Hefenbrock,Mehdi B Tahoori,Jasmin Aghassi-Hagmann

doi:10.1038/s41598-021-88396-0

Abstract

Emerging applications in soft robotics, wearables, smart consumer products or IoT-devices benefit from soft materials, flexible substrates in conjunction with electronic functionality. Due to high production costs and conformity restrictions, rigid silicon technologies do not meet application requirements in these new domains. However, whenever signal processing becomes too comprehensive, silicon technology must be used for the high-performance computing unit. At the same time, designing everything in flexible or printed electronics using conventional digital logic is not feasible yet due to the limitations of printed technologies in terms of performance, power and integration density. We propose to rather use the strengths of neuromorphic computing architectures consisting in their homogeneous topologies, few building blocks and analog signal processing to be mapped to an inkjet-printed hardware architecture. It has remained a challenge to demonstrate non-linear elements besides weighted aggregation. We demonstrate in this work printed hardware building blocks such as inverter-based comprehensive weight representation and resistive crossbars as well as printed transistor-based activation functions. In addition, we present a learning algorithm developed to train the proposed printed NCS architecture based on specific requirements and constraints of the technology.

Highlights

Emerging applications in soft robotics, wearables, smart consumer products or Internet of things (IoT)-devices benefit from soft materials, flexible substrates in conjunction with electronic functionality
We address the aforementioned shortcomings and demonstrate that all fundamental building blocks for printed artificial neural networks (ANN) can be fabricated within the same process and materials, such as MAC operation, non-linear activation functions and comprehensive weights representation
We will present in the following, designs and hardware prototypes of printed neuromorphic computing system (NCS) components at the circuit level, which can be used as basic building blocks to realize arbitrary large ANN architectures in Printed Electronics (PE)

Summary

Introduction

Emerging applications in soft robotics, wearables, smart consumer products or IoT-devices benefit from soft materials, flexible substrates in conjunction with electronic functionality. We address the aforementioned shortcomings and demonstrate that all fundamental building blocks for printed ANNs can be fabricated within the same process and materials, such as MAC operation, non-linear activation functions and comprehensive weights representation.

Results

Conclusion