Preface
Conventional Von Neumann computing architectures based on CMOS technology
are currently facing challenges termed “the heat and memory wall” in addition to
the advent of Moore’s Law slowdown [1–3]. Von Neumann bottleneck originates in
particular from the speed limitations due to the constant data movements between
the memory (e.g., the Random Access Memory—RAM) and the Central Processing
Unit (CPU), since RAM and CPU have physically distinct locations. Going beyond
CMOS and overcoming Von Neumann restrictions are long-term visions aimed at
developing completely new nanoscale components with unconventional functions
and dynamics that are capable of outperforming similar CMOS implementations to
sustain the growth of the electronics industry at the end of Moore’s Law.
The memristor (a shorthand for memory-resistor) is one of the most promising
candidate information processing devices for beyond CMOS and more than Moore
semiconductor technology. The memristor has been theoretically envisioned by L.
O. Chua in 1971 [4] as the fourth basic passive circuit element, in addition to the
resistor, inductor, and capacitor, using an axiomatic approach on device modeling
and symmetry arguments on the basic electric quantities. A memristor is a statedependent resistor, where the resistance (also named memristance) is not fixed but
rather depends on the history of the voltage or current. A memristor is endowed
with a number of new peculiar features that are not shared by the other basic circuit
elements.
For a long time, the memristor remained basically an object of academic interest
since no passive physical device was known behaving as a memristor. Almost four
decades after the publication of the seminal paper [4], the researchers at Hewlett
and Packard headed by R. Stanley Williams first announced and presented to the
world a nanoscale device displaying memristive features [5]. This work has sprung
a huge worldwide cross-disciplinary interest on memristor and its applications
ranging from tunable electronics, neuromorphic/in-memory computing, biosensors,
data storage, and complex nonlinear systems [6–8].
On-chip memory, biologically inspired computing and in-memory computing,
i.e., the integration of storage and computation in the same physical location [9], are
categories that are expected to significantly benefit from memristor developments.
xiii
Conventional Von Neumann computing architectures based on CMOS technology
are currently facing challenges termed “the heat and memory wall” in addition to
the advent of Moore’s Law slowdown [1–3]. Von Neumann bottleneck originates in
particular from the speed limitations due to the constant data movements between
the memory (e.g., the Random Access Memory—RAM) and the Central Processing
Unit (CPU), since RAM and CPU have physically distinct locations. Going beyond
CMOS and overcoming Von Neumann restrictions are long-term visions aimed at
developing completely new nanoscale components with unconventional functions
and dynamics that are capable of outperforming similar CMOS implementations to
sustain the growth of the electronics industry at the end of Moore’s Law.
The memristor (a shorthand for memory-resistor) is one of the most promising
candidate information processing devices for beyond CMOS and more than Moore
semiconductor technology. The memristor has been theoretically envisioned by L.
O. Chua in 1971 [4] as the fourth basic passive circuit element, in addition to the
resistor, inductor, and capacitor, using an axiomatic approach on device modeling
and symmetry arguments on the basic electric quantities. A memristor is a statedependent resistor, where the resistance (also named memristance) is not fixed but
rather depends on the history of the voltage or current. A memristor is endowed
with a number of new peculiar features that are not shared by the other basic circuit
elements.
For a long time, the memristor remained basically an object of academic interest
since no passive physical device was known behaving as a memristor. Almost four
decades after the publication of the seminal paper [4], the researchers at Hewlett
and Packard headed by R. Stanley Williams first announced and presented to the
world a nanoscale device displaying memristive features [5]. This work has sprung
a huge worldwide cross-disciplinary interest on memristor and its applications
ranging from tunable electronics, neuromorphic/in-memory computing, biosensors,
data storage, and complex nonlinear systems [6–8].
On-chip memory, biologically inspired computing and in-memory computing,
i.e., the integration of storage and computation in the same physical location [9], are
categories that are expected to significantly benefit from memristor developments.
xiii