Skip to main navigation Skip to search Skip to main content

Reliability-Ensured and Fast (< 100 ns) Analog Synpase for Training Accelerators: All-Sputtered HfOy/HfOxRRAM

  • Kyungpook National University

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

2 Scopus citations

Abstract

We reveal that sputtered HfOx, which has been often neglected in resistive random access memory (RRAM) design, can be an alternative for highly reliable and fast (< 100 ns) analog synapses to accelerate training algorithms. The sub-stoichiometric HfOx, enables rapid current update owing to the abundant oxygen vacancies (V0). Additional HfOy layer with increased stoichiometry makes uniformly distributed Vo profile across the entire HfOy/HfOxRRAM stack, enhancing reliability in dynamic synaptic operation, where multiple pulses are continuously addressed. This not only results in robust device-to-device/cycle-to-cycle uniformity, but also ensures multilevel retention, which is further validated by real-time transient current analysis.

Original languageEnglish
Title of host publication2025 IEEE International Reliability Physics Symposium, IRPS 2025 - Proceedings
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9798331504779
DOIs
StatePublished - 2025
Event2025 IEEE International Reliability Physics Symposium, IRPS 2025 - Monterey, United States
Duration: 30 Mar 20253 Apr 2025

Publication series

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

Conference

Conference2025 IEEE International Reliability Physics Symposium, IRPS 2025
Country/TerritoryUnited States
CityMonterey
Period30/03/253/04/25

Keywords

  • Neuromorphic computing
  • RRAM
  • synaptic device

Fingerprint

Dive into the research topics of 'Reliability-Ensured and Fast (< 100 ns) Analog Synpase for Training Accelerators: All-Sputtered HfOy/HfOxRRAM'. Together they form a unique fingerprint.

Cite this