Open AccessModelling of wire resistance effect in PCM‐based nanocrossbar memory
Author(s) -
ElHassan Nemat H.,
Kumar Nandha Thulasiraman,
Almurib Haider Abbas F.
Publication year - 2016
Publication title -
the journal of engineering
Language(s) - English
Resource type - Journals
ISSN - 2051-3305
DOI - 10.1049/joe.2016.0212
Subject(s) - materials science , resistive touchscreen , electrical resistivity and conductivity , power (physics) , contact resistance , phase change memory , optoelectronics , composite material , electrical engineering , physics , engineering , thermodynamics , layer (electronics)
At nanoscale grain boundaries and surface scattering effects lead to an increase of connecting wires electrical resistivity with decreasing wire dimensions. This increase of resistivity leads to significant power loss across connecting wires in nanocrossbars. In this study, the resistance of connecting wire as a function of material properties and feature size is calculated. Then the effect of the connecting wires resistance in phase change memory (PCM) performance in PCM‐based passive nanocrossbar was evaluated. The performance metrics tested are: programmed resistance levels, programming duration, and energy consumption. Based on the simulation results, it was found that the power consumed in connecting wires decreases the power supplied to PCM cells. This reduction in power results in higher programmed low resistive state ( R ON ). The effect of connecting wire resistance on PCM performance is studied as a function of the wire size, cell position on the nanocrossbar, and nanocrossbar size. Simulation results showed that the programmed R ON is inversely proportional to feature size. Moreover, it increases up to almost 40%, with decreasing feature size to 40 nm. Moreover, programmed R ON increases proportionally with increasing nanocrossbar size. Moreover, R OFF / R ON ratio drops almost 90% of targeted ratio at 1 kbit nanocrossbars. Furthermore, cells closer to supply sources are the least affected by wire resistance, while cells furthest from supply are the most affected. Finally, at the end of this study two methods are suggested to resolve the programmed R ON reliability issue caused by energy drop across connecting wires.