loading page

Design of Variation-Tolerant 1F-1T Memory Array for Neuromorphic Computing
  • +11
  • Masud S K Rana ,
  • Sunanda Thunder ,
  • Franz Müller ,
  • Nellie Laleni ,
  • Yannick Raffel ,
  • Maximilian Lederer ,
  • Luca Pirro ,
  • Talha Chohan ,
  • Jing-Hua Hsuen ,
  • Tian-Li Wu ,
  • Konrad Seidel ,
  • Thomas Kaempfe ,
  • Sourav De ,
  • Bhaswar Chakrabarti
Masud S K Rana
Author Profile
Sunanda Thunder
Author Profile
Franz Müller
Author Profile
Nellie Laleni
Author Profile
Yannick Raffel
Author Profile
Maximilian Lederer
Author Profile
Luca Pirro
Author Profile
Talha Chohan
Author Profile
Jing-Hua Hsuen
Author Profile
Tian-Li Wu
Author Profile
Konrad Seidel
Author Profile
Thomas Kaempfe
Author Profile
Sourav De
Author Profile
Bhaswar Chakrabarti
Indian Institute of Technology Madras

Corresponding Author:[email protected]

Author Profile

Abstract

This letter proposes a memory cell, denoted by 1F-1T, consisting of a ferroelectric field-effect transistor (Fe-FET) cascoded with another current-limiting transistor (T). The transistor reduces the impact of drain current (Id) variations by limiting the on-state current in FeFET, denoted by 1F. We have fabricated 28nm high-k-meta-gate (HKMG) based FeFETs, and the experimental data is used to model and simulate single-cell and memory arrays. The simulation shows significant improvement in bit-line (BL) current (IBL ) variation for 1F-1T memory array. Finally, the system-level neuromorphic simulation with 1F-1T synapses shows an inference accuracy of 97.6% for MNIST hand-written digits with multi-layer perceptron (MLP) neural networks.