Chapter 10
Extended Memristor Devices
Memristor devices embedded in pioneering future computing platforms are believed
to represent one of the most promising key-enabling technologies for the treatment
of massive amount of data. The ideal memristor, a two-terminal circuit element
described by a nonlinear relationship between the time integrals of current and
voltage (i.e., the current and voltage momenta, a.k.a. charge and flux), was
theoretically envisioned by Chua back in 1971 in the seminal paper [1]. In 2008,
some main peculiar features of the memristor were identified in a nano-scale
film based on titanium dioxide by a team of Hewlett Packard researchers headed
by S. Williams [2] (Chap. 2). After this discovery, the scientific community has
focused the interest on the experimental observation of some aspects of memristor
behavior in other nano-structures (Chap. 2). Memristor nano-devices typically adopt
a metal/insulator/metal (MIM) structure. The electrode materials and the switching layer are carefully designed so as to obtain desirable programming voltage,
memristance on/off ratio, power consumption, and device variation that enable fast,
low-power yet reliable data processing.
Real memristor devices may exhibit a number of peculiar features that are in
general not displayed by ideal memristors (cf. Chap. 2). For instance, different from
an ideal memristor, a real memristor device may be volatile, it may have voltage
thresholds and asymmetries in the hysteresis loops in the (v, i)-domain, it may
display boundary effects due to the interface between the electrodes and the oxide,
or it may show a dependence of memductance on different (nonelectrical) physical
state variables (e.g., temperature, ion concentration, etc.) [3–6]. Roughly speaking,
memristor devices can be modeled by a generic nonlinear dynamical input–output
system just satisfying the fundamental zero-crossing property, i.e., the output (e.g.,
the current) vanishes when the input (e.g., the voltage) is zero. As a consequence,
pinched hysteresis loops in the (v, i)-domain are observed when memristor devices
are subject to a (zero-mean bipolar) periodic input. Actually, such systems satisfying
a zero-crossing property fall into the class of the so-called memristive devices and
systems introduced by Chua and Kang in [7], that have been discussed in Sect. 2.3
© Springer Nature Switzerland AG 2021
F. Corinto et al., Nonlinear Circuits and Systems with Memristors,
https://doi.org/10.1007/978-3-030-55651-8_10
373
Extended Memristor Devices
Memristor devices embedded in pioneering future computing platforms are believed
to represent one of the most promising key-enabling technologies for the treatment
of massive amount of data. The ideal memristor, a two-terminal circuit element
described by a nonlinear relationship between the time integrals of current and
voltage (i.e., the current and voltage momenta, a.k.a. charge and flux), was
theoretically envisioned by Chua back in 1971 in the seminal paper [1]. In 2008,
some main peculiar features of the memristor were identified in a nano-scale
film based on titanium dioxide by a team of Hewlett Packard researchers headed
by S. Williams [2] (Chap. 2). After this discovery, the scientific community has
focused the interest on the experimental observation of some aspects of memristor
behavior in other nano-structures (Chap. 2). Memristor nano-devices typically adopt
a metal/insulator/metal (MIM) structure. The electrode materials and the switching layer are carefully designed so as to obtain desirable programming voltage,
memristance on/off ratio, power consumption, and device variation that enable fast,
low-power yet reliable data processing.
Real memristor devices may exhibit a number of peculiar features that are in
general not displayed by ideal memristors (cf. Chap. 2). For instance, different from
an ideal memristor, a real memristor device may be volatile, it may have voltage
thresholds and asymmetries in the hysteresis loops in the (v, i)-domain, it may
display boundary effects due to the interface between the electrodes and the oxide,
or it may show a dependence of memductance on different (nonelectrical) physical
state variables (e.g., temperature, ion concentration, etc.) [3–6]. Roughly speaking,
memristor devices can be modeled by a generic nonlinear dynamical input–output
system just satisfying the fundamental zero-crossing property, i.e., the output (e.g.,
the current) vanishes when the input (e.g., the voltage) is zero. As a consequence,
pinched hysteresis loops in the (v, i)-domain are observed when memristor devices
are subject to a (zero-mean bipolar) periodic input. Actually, such systems satisfying
a zero-crossing property fall into the class of the so-called memristive devices and
systems introduced by Chua and Kang in [7], that have been discussed in Sect. 2.3
© Springer Nature Switzerland AG 2021
F. Corinto et al., Nonlinear Circuits and Systems with Memristors,
https://doi.org/10.1007/978-3-030-55651-8_10
373