Printed Machine Learning Classifiers

doi:10.1109/MICRO50266.2020.00019

Return

Printed Machine Learning Classifiers

2020-10-01

PRE

Muhammad Husnain Mubarik *

Dennis D. Weller

Nathaniel Bleier

Matthew Tomei

Jasmin Aghassi‐Hagmann

Mehdi B. Tahoori

Rakesh Kumar

DOI：10.1109/MICRO50266.2020.00019

Original

Original request for help

Save

Abstract

En 中文

A large number of application domains have requirements on cost, conformity, and non-toxicity that silicon-based computing systems cannot meet, but that may be met by printed electronics. For several of these domains, a typical computational task to be performed is classification. In this work, we explore the hardware cost of inference engines for popular classification algorithms (Multi-Layer Perceptrons, Support Vector Machines (SVMs), Logistic Regression, Random Forests and Binary Decision Trees) in EGT and CNT-TFT printed technologies and determine that Decision Trees and SVMs provide a good balance between accuracy and cost. We evaluate conventional Decision Tree and SVM architectures in these technologies and conclude that their area and power overhead must be reduced. We explore, through SPICE and gate-level hardware simulations and multiple working prototypes, several classifier architectures that exploit the unique cost and implementation tradeoffs in printed technologies - a) Bespoke printed classifers that are customized to a model generated for a given application using specific training datasets, b) Lookup-based printed classifiers where key hardware computations are replaced by lookup tables, and c) Analog printed classifiers where some classifier components are replaced by their analog equivalents. Our evaluations show that bespoke implementation of EGT printed Decision Trees has 48.9x lower area (average) and 75.6x lower power (average) than their conventional equivalents; corresponding benefits for bespoke SVMs are 12.8x and 12.7x respectively. Lookup-based Decision Trees outperform their non-lookup bespoke equivalents by 38% and 70%; lookup-based SVMs are better by 8% and 0.6%. Analog printed Decision Trees provide 437 x area and 27x power benefits over digital bespoke counterparts; analog SVMs yield 490x area and 12x power improvements. Our results and prototypes demonstrate feasibility of fabricating and deploying battery and self-powered printed classifiers in the application domains of interest.

Keywords:

printed electronics

machine learning

AI Summary

Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.