A low-power rram memory block for embedded, multi-level weight and bias storage in artificial neural networks

Stefan Pechmann, Timo Mai, Julian Potschka, Daniel Reiser, Peter Reichel, Marco Breiling, Marc Reichenbach, Amelie Hagelauer

Research output: Contribution to journalArticlepeer-review

6 Scopus citations

Abstract

Pattern recognition as a computing task is very well suited for machine learning algorithms utilizing artificial neural networks (ANNs). Computing systems using ANNs usually require some sort of data storage to store the weights and bias values for the processing elements of the individual neurons. This paper introduces a memory block using resistive memory cells (RRAM) to realize this weight and bias storage in an embedded and distributed way while also offering programming and multi-level ability. By implementing power gating, overall power consumption is decreased significantly without data loss by taking advantage of the non-volatility of the RRAM technology. Due to the versatility of the peripheral circuitry, the presented memory concept can be adapted to different applications and RRAM technologies.

Original languageEnglish
Article number1277
JournalMicromachines
Volume12
Issue number11
DOIs
StatePublished - Nov 2021

Keywords

  • ANN
  • Embedded memory
  • Low-power
  • Memory block
  • Multi-level
  • RRAM

Fingerprint

Dive into the research topics of 'A low-power rram memory block for embedded, multi-level weight and bias storage in artificial neural networks'. Together they form a unique fingerprint.

Cite this