Memristors for Energy‐Efficient New Computing Paradigms

Abstract

In this Review, memristors are examined from the frameworks of both von Neumann and neuromorphic computing architectures. For the former, a new logic computational process based on the material implication is discussed. It consists of several memristors which play roles of combined logic processor and memory, called stateful logic circuit. In this circuit confi guration, the logic process fl ows primarily along a time dimension, whereas in current von Neumann computers it occurs along a spatial dimension. In the stateful logic computation scheme, the energy required for the data transfer between the logic and memory chips can be saved. The non-volatile memory in this circuit also saves the energy required for the data refresh. Neuromorphic (cognitive) computing refers to a computing paradigm that mimics the human brain. Currently, the neuromorphic or cognitive computing mainly relies on the software emulation of several brain functionalities, such as image and voice recognition utilizing the recently highlighted deep learning algorithm. However, the human brain typically consumes ≈10-20 Watts for selected "human-like" tasks, which can be currently mimicked by a supercomputer with power consumption of several tens of kilo-to megawatts. Therefore, hardware implementation of such brain functionality must be eventually sought for power-effi cient computation. Several fundamental ideas for utilizing the memristors and their recent progresses in these regards are reviewed. Finally, material and processing issues are dealt with, which is followed by the conclusion and outlook of the fi eld. These technical improvements will substantially decrease the energy consumption for futuristic information technology.

Research Update: Fast and tunable nanoionics in vertically aligned nanostructured films

“…A decrease of the film thickness into the nanoscale regime has boosted the development of ionic memory devices. [19][20][21][22][23][24][25][26]30,35 As shown in Figure 1(c), the device structures are usually composed of oxide thin films (or semiconductor thin films) with either novel metal (e.g., Au and Pt) electrodes or mobile metal (e.g., Ag or Cu) electrodes. However, most as-grown films are not sufficient to display electric-field-induced RS phenomena.…”

“…Next-generation memory devices based on resistive switching (RS) phenomena have been demonstrated using ionic motion in semiconductors. [19][20][21][22][23][24][25][26] When an electric field is applied, ions move toward one electrode, and accumulate and form conducting channels, resulting in a low-resistive state of the device. When an electric field with opposite polarity is applied, the ions in the channels move back to the other electrode; therefore, the channels are dissolved and ruptured, resulting in a high-resistive state of the device.…”

What Will Come After V‐NAND—Vertical Resistive Switching Memory?

“…[56][57][58] The reversible switching of the phase and the accompanying optical and electrical property change were first reported by Ovshinsky in 1968, but the focus then was on the optical property change. [5,60] In this regard, such a return of Intel to the memory business may have been a natural consequence of its stagnant central processing unit (CPU) business, but the conventional DRAM and NAND flash markets are dominated by the companies in South Korea and Japan. This is a natural consequence of the recent change in the semiconductor market trend.…”

Resistive Memory‐Based In‐Memory Computing: From Device and Large‐Scale Integration System Perspectives

“…[47] Following the classical Boolean logic to construct computing automata, logic gate implementation with memory devices is the first attempt to exceed the function of memory and head for computation. [29,32,33] In 2012, Hu et al described how to conduct VMM on an RRAM crossbar array based on Kirchhoff's law. [55,56] With parallel or serial connection setups, sensing voltage is applied to the RRAM devices with data stored as resistances.…”

“…[23][24][25][26] Racetrack memory is known for its extremely high density (20 nm-wide nanowire) and the sequential access along tracks. [32] Tsai et al introduced the principal applications of analog memory devices in building deep learning accelerators, including prospective candidates of resistive, capacitive, and photonics devices. [18] RRAM offers versatility, including high resistivity (MΩ order of magnitude), the support of 3D integration, stochastic programming, and multilevel cell (up to 6 bits).…”