Characterization and Mitigation of Relaxation Effects on Multi-level RRAM based In-Memory Computing

Wangxin He, Wonbo Shim, Shihui Yin, Xiaoyu Sun, Deliang Fan, Shimeng Yu, Jae Sun Seo

Research output: Chapter in Book/Report/Conference proceedingConference contribution

8 Scopus citations

Abstract

In this paper, we investigate the relaxation effects on multi-level resistive random access memory (RRAM) based in-memory computing (IMC) for deep neural network (DNN) inference. We characterized 2-bit-per-cell RRAM IMC prototypes and measured the relaxation effects over 100 hours on multiple 8 kb test chips, where the relaxation is found to be most severe in the two intermediate states. We incorporated the experimental data into SPICE simulation and software DNN inference, showing DNN accuracy for CIFAR-10 dataset could degrade from 87.35% to 11.58% after 144 hours. To recover the largely degraded accuracy, mitigation schemes are proposed: 1) at the circuit level, the reference voltage for RRAM IMC could be calibrated after 80 hours when the relaxation is saturated. 2) At the algorithm level, the weights are trained with lower percentages to be quantized to the two intermediate states. With both schemes applied, the accuracy could be recovered to 87.32 % for long-term stability.

Original languageEnglish (US)
Title of host publication2021 IEEE International Reliability Physics Symposium, IRPS 2021 - Proceedings
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781728168937
DOIs
StatePublished - Mar 2021
Event2021 IEEE International Reliability Physics Symposium, IRPS 2021 - Virtual, Monterey, United States
Duration: Mar 21 2021Mar 24 2021

Publication series

NameIEEE International Reliability Physics Symposium Proceedings
Volume2021-March
ISSN (Print)1541-7026

Conference

Conference2021 IEEE International Reliability Physics Symposium, IRPS 2021
Country/TerritoryUnited States
CityVirtual, Monterey
Period3/21/213/24/21

Keywords

  • RRAM
  • deep neural network
  • in-memory computing
  • multi-level cell
  • relaxation effect

ASJC Scopus subject areas

  • General Engineering

Fingerprint

Dive into the research topics of 'Characterization and Mitigation of Relaxation Effects on Multi-level RRAM based In-Memory Computing'. Together they form a unique fingerprint.

Cite this