Exploiting Oxide Based Resistive RAM Variability for Bayesian Neural Network Hardware Design

Akul Malhotra, Sen Lu, Kezhou Yang, Abhronil Sengupta

Research output: Contribution to journalArticlepeer-review

Abstract

Uncertainty plays a key role in real-time machine learning. As a significant shift from standard deep networks, which does not consider any uncertainty formulation during its training or inference, Bayesian deep networks are being currently investigated where the network is envisaged as an ensemble of plausible models learnt by the Bayes' formulation in response to uncertainties in sensory data. Bayesian deep networks consider each synaptic weight as a sample drawn from a probability distribution with learnt mean and variance. This letter elaborates on a hardware design that exploits cycle-to-cycle variability of oxide based Resistive Random Access Memories (RRAMs) as a means to realize such a probabilistic sampling function, instead of viewing it as a disadvantage.

Original languageEnglish (US)
Article number9050663
Pages (from-to)328-331
Number of pages4
JournalIEEE Transactions on Nanotechnology
Volume19
DOIs
StatePublished - 2020

All Science Journal Classification (ASJC) codes

  • Computer Science Applications
  • Electrical and Electronic Engineering

Fingerprint Dive into the research topics of 'Exploiting Oxide Based Resistive RAM Variability for Bayesian Neural Network Hardware Design'. Together they form a unique fingerprint.

Cite this