Foreword by Sung Mo (Steve) Kang
Nonlinear Circuits and Systems with Memristors is a timely contribution to the field
of nanoelectronic circuits and systems. Personally, the contents of this book are dear
to me, especially since I was one of the early graduate students of Professor Leon
Chua at the UC Berkeley who at that time introduced memristors and memristive
systems. The seminal paper on memristor by Leon Chua was published in 1971,
followed by the Chua and Kang paper on memristive devices and systems in 1976. In
microelectronics industry, CMOS Very Large-Scale Integrated (VLSI) circuits have
been the dominant workhorse, the development of which has followed the Moore’s
law. However, as the downscaling faced its limitations and because of the volatility
of charge storage in ultrasmall capacitors in VLSI chips, nonvolatile resistance has
become critically important as a new state variable. The demands for Resistive
RAMs (RRAMs) and other memory devices such as MRAM, PCRAM, STT-RAM,
which do not depend on charge storage, have increased. Analog computing has also
become of increasing importance for ultralow energy computing as in neuromorphic
computing. Almost four decades later, when Stan Williams and his associates in HP
announced nanoscale memristors in the May 2008 Nature paper, a new epoch for
integrated memristor circuits was established. HP’s memristor was the first solidstate realization of the “two-terminal memristor.” It was my honor and pleasure to
organize the first symposium on memristors and memristive systems in November
2008 under sponsorship of NSF and HP at the Berkeley campus, the fountain
of the memristor research. Since then, memristor electronics has been pursued
globally at a phenomenal growth rate. In particular, the application of crossbar
arrays of memristors for analog neuromorphic computing has been pursued actively
as one of the most promising approaches to the mimicry of brain functions. Twoterminal memristors are used in crossbar arrays with programmable memductances
as learning weights. Coincidentally in their Proceedings of the IEEE paper (1976),
Chua and Kang identified the conductive channels of the Hodgkin–Huxley model
for neuromorphic signal (action potential) generation to be memristive. A few years
ago, after giving a seminar on memristors at the Rowan University in New Jersey,
its faculty members asked me when a textbook would be published on the subject
matter for education of electronic circuits including memristors. Although this book
vii
Nonlinear Circuits and Systems with Memristors is a timely contribution to the field
of nanoelectronic circuits and systems. Personally, the contents of this book are dear
to me, especially since I was one of the early graduate students of Professor Leon
Chua at the UC Berkeley who at that time introduced memristors and memristive
systems. The seminal paper on memristor by Leon Chua was published in 1971,
followed by the Chua and Kang paper on memristive devices and systems in 1976. In
microelectronics industry, CMOS Very Large-Scale Integrated (VLSI) circuits have
been the dominant workhorse, the development of which has followed the Moore’s
law. However, as the downscaling faced its limitations and because of the volatility
of charge storage in ultrasmall capacitors in VLSI chips, nonvolatile resistance has
become critically important as a new state variable. The demands for Resistive
RAMs (RRAMs) and other memory devices such as MRAM, PCRAM, STT-RAM,
which do not depend on charge storage, have increased. Analog computing has also
become of increasing importance for ultralow energy computing as in neuromorphic
computing. Almost four decades later, when Stan Williams and his associates in HP
announced nanoscale memristors in the May 2008 Nature paper, a new epoch for
integrated memristor circuits was established. HP’s memristor was the first solidstate realization of the “two-terminal memristor.” It was my honor and pleasure to
organize the first symposium on memristors and memristive systems in November
2008 under sponsorship of NSF and HP at the Berkeley campus, the fountain
of the memristor research. Since then, memristor electronics has been pursued
globally at a phenomenal growth rate. In particular, the application of crossbar
arrays of memristors for analog neuromorphic computing has been pursued actively
as one of the most promising approaches to the mimicry of brain functions. Twoterminal memristors are used in crossbar arrays with programmable memductances
as learning weights. Coincidentally in their Proceedings of the IEEE paper (1976),
Chua and Kang identified the conductive channels of the Hodgkin–Huxley model
for neuromorphic signal (action potential) generation to be memristive. A few years
ago, after giving a seminar on memristors at the Rowan University in New Jersey,
its faculty members asked me when a textbook would be published on the subject
matter for education of electronic circuits including memristors. Although this book
vii