Reliability and Performance Analysis of Logic-in-Memory Based Binarized Neural Networks
- 23 April 2021
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Device and Materials Reliability
- Vol. 21 (2) , 183-191
Abstract
Resistive Random access memory (RRAM) devices together with the material implication (IMPLY) logic are a promising computing scheme for realizing energy efficient reconfigurable computing hardware for edge computing applications. This approach has been recently shown to enable the in-memory implementation of Binarized Neural Networks. However, an accurate analysis of the performance achieved on a real classification task are still missing. In this work, we train and estimate the performance of an IMPLY-based implementation of a multilayer perceptron (MLP) BNN and highlight its main reliability challenges by using a physics-based RRAM compact model calibrated on three RRAM technologies from the literature. We then show how the smart IMPLY (SIMPLY) architecture solves the reliability issues of conventional IMPLY architectures and compare its performance with respect to conventional solutions considering different parallelization degree. The worst-case energy estimates for an inference task performed on the trained network, show that the SIMPLY implementation results in a >46 energy-delay-product (EDP) improvement with respect to a conventional low-power embedded system implementation.Keywords
This publication has 20 references indexed in Scilit:
- Multilevel conductance switching of a HfO2 RRAM array induced by controlled filament for neuromorphic applications2016
- Improved Synaptic Behavior Under Identical Pulses Using AlOx/HfO2Bilayer RRAM Array for Neuromorphic SystemsIEEE Electron Device Letters, 2016
- Bipolar Resistive RAM Based on ${\rm HfO}_{2}$ : Physics, Compact Modeling, and Variability ControlIEEE Journal on Emerging and Selected Topics in Circuits and Systems, 2016
- Very Low-Programming-Current RRAM With Self-Rectifying CharacteristicsIEEE Electron Device Letters, 2016
- A neuromorphic visual system using RRAM synaptic devices with Sub-pJ energy and tolerance to variability: Experimental characterization and large-scale modeling2012
- $\hbox{HfO}_{x}/\hbox{TiO}_{x}/\hbox{HfO}_{x}/ \hbox{TiO}_{x}$ Multilayer-Based Forming-Free RRAM Devices With Excellent UniformityIEEE Electron Device Letters, 2011
- ‘Memristive’ switches enable ‘stateful’ logic operations via material implicationNature, 2010
- Two memristors suffice to compute all Boolean functionsElectronics Letters, 2010
- FreePDK: An Open-Source Variation-Aware Design Kit2007
- Gradient-based learning applied to document recognitionProceedings of the IEEE, 1998