Weight update in ferroelectric memristors with identical and non-identical pulses
Suzanne Lancaster, Maximilien Remillieux, Moritz Engl, Viktor, Havel, Claudia Silva, Xuetao Wang, Thomas Mikolajick, Stefan, Slesazeck

TL;DR
This paper explores methods for weight updates in ferroelectric tunnel junctions using identical and non-identical pulses, demonstrating improved linearity and control through experimental and simulation techniques for potential low-power neuromorphic applications.
Contribution
It introduces a novel approach to achieve linear weight updates with identical pulses at long delays by limiting switching current, supported by experimental and simulation results.
Findings
Achieved non-linear weight update with identical pulses using a series resistor.
Maximum linearity of 86% in a 1T1C cell with current limiting.
Enhanced linearity to over 93% by engineering device switching characteristics.
Abstract
Ferroelectric tunnel junctions (FTJs) are a class of memristor which promise low-power, scalable, field-driven analog operation. In order to harness their full potential, operation with identical pulses is targeted. In this paper, several weight update schemes for FTJs are investigated, using either non-identical or identical pulses, and with time delays between the pulses ranging from 1 us to 10 s. Experimentally, a method for achieving non-linear weight update with identical pulses at long programming delays is demonstrated by limiting the switching current via a series resistor. Simulations show that this concept can be expanded to achieve weight update in a 1T1C cell by limiting the switching current through a transistor operating in sub-threshold or saturation mode. This leads to a maximum linearity in the weight update of 86% for a dynamic range (maximum switched polarization) of…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsAdvanced Memory and Neural Computing · Phase-change materials and chalcogenides · Semiconductor materials and interfaces
