Hardware-aware Training Techniques for Improving Robustness of Ex-Situ Neural Network Transfer onto Passive TiO2 ReRAM Crossbars

by   Philippe Drolet, et al.

Passive resistive random access memory (ReRAM) crossbar arrays, a promising emerging technology used for analog matrix-vector multiplications, are far superior to their active (1T1R) counterparts in terms of the integration density. However, current transfers of neural network weights into the conductance state of the memory devices in the crossbar architecture are accompanied by significant losses in precision due to hardware variabilities such as sneak path currents, biasing scheme effects and conductance tuning imprecision. In this work, training approaches that adapt techniques such as dropout, the reparametrization trick and regularization to TiO2 crossbar variabilities are proposed in order to generate models that are better adapted to their hardware transfers. The viability of this approach is demonstrated by comparing the outputs and precision of the proposed hardware-aware network with those of a regular fully connected network over a few thousand weight transfers using the half moons dataset in a simulation based on experimental data. For the neural network trained using the proposed hardware-aware method, 79.5 the test set's data points can be classified with an accuracy of 95 while only 18.5 accuracy by the regularly trained neural network.


page 2

page 8

page 13


Deep Neural Network Optimized to Resistive Memory with Nonlinear Current-Voltage Characteristics

Artificial Neural Network computation relies on intensive vector-matrix ...

High-Accuracy Inference in Neuromorphic Circuits using Hardware-Aware Training

Neuromorphic Multiply-And-Accumulate (MAC) circuits utilizing synaptic w...

Implementation of a Binary Neural Network on a Passive Array of Magnetic Tunnel Junctions

The increasing scale of neural networks and their growing application sp...

Methodology for Realizing VMM with Binary RRAM Arrays: Experimental Demonstration of Binarized-ADALINE Using OxRAM Crossbar

In this paper, we present an efficient hardware mapping methodology for ...

On-chip learning for domain wall synapse based Fully Connected Neural Network

Spintronic devices are considered as promising candidates in implementin...

Device Modeling Bias in ReRAM-based Neural Network Simulations

Data-driven modeling approaches such as jump tables are promising techni...

Cognitive Memory Network

A resistive memory network that has no crossover wiring is proposed to o...

Please sign up or login with your details

Forgot password? Click here to reset