Benchmarking the performance of heterogeneous stacked RRAM with CFETSRAM and MRAM for deep neural network application amidst variation and noise

Parthasarathi Pal, Sunanda Thunder, Min Jung Tsai, Po-Tsang Huang*, Yeong Her Wang

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

8 Scopus citations

Abstract

In this article we demonstrate and compare the performance of 32nm technology node compatible high-K and low-K stacked RRAM with CFET-SRAM and MRAM for binary deep neural network. We have fabricated heterogenous stacked RRAM with Sidoped Al2O3 and Ta2O5 as stacked layer for synaptic memory application. The device demonstrated an exorbitant on/off ratio ~ 4.2 x 103 with an ultra-low variation (σ ~ 6E-07 S). We have trained the neural network with 97.11% accuracy as baseline and observed the impact of conductance variation and read noise variation. We have also benchmarked the performance of our device with CFET-SRAM and MRAM technologies from other works and observed superior performance of our devices in terms of accuracy.

Original languageEnglish
Title of host publicationVLSI-TSA 2021 - 2021 International Symposium on VLSI Technology, Systems and Applications, Proceedings
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages1-2
Number of pages2
ISBN (Electronic)9781665419345
DOIs
StatePublished - 19 Apr 2021
Event2021 International Symposium on VLSI Technology, Systems and Applications, VLSI-TSA 2021 - Hsinchu, Taiwan
Duration: 19 Apr 202122 Apr 2021

Publication series

NameVLSI-TSA 2021 - 2021 International Symposium on VLSI Technology, Systems and Applications, Proceedings

Conference

Conference2021 International Symposium on VLSI Technology, Systems and Applications, VLSI-TSA 2021
Country/TerritoryTaiwan
CityHsinchu
Period19/04/2122/04/21

Keywords

  • Bipolar resistive switching
  • Neural network
  • Recognition accuracy
  • ReRAM
  • Synaptic simulation

Fingerprint

Dive into the research topics of 'Benchmarking the performance of heterogeneous stacked RRAM with CFETSRAM and MRAM for deep neural network application amidst variation and noise'. Together they form a unique fingerprint.

Cite this