Resistance switching (RS) devices have potential to offer computing and memory function. A new computer unit is built of RS array, where processing storing information occur on same devices. states stored in located arbitrary positions array can be performed various nonvolatile logic operations. Logic functions reconfigured by altering trigger signals.

The exponential growth of various complex images is putting tremendous pressure on storage systems. Here, we propose a memristor-based system with an integrated near-storage in-memory computing-based convolutional autoencoder compression network to boost the energy efficiency and speed image compression/retrieval improve density. We adopt 4-bit memristor arrays experimentally demonstrate functions system. step-by-step quantization aware training scheme equivalent transformation for transpose...

Reservoir computing is a powerful neural network-based paradigm for spatiotemporal signal processing. Recently, physical reservoirs have been explored based on various electronic devices with outstanding efficiency. However, the inflexible temporal dynamics of these posed fundamental restrictions in processing signals timescales. Here, we fabricated thin-film transistors controllable dynamics, which can be easily tuned electrical operation and showed excellent cycle-to-cycle uniformity....

A physics-based compact model of metal-oxide-based resistive-switching random access memory (RRAM) cell under dc and ac operation modes is presented. In this model, the conductive filament evolution corresponding to resistive switching process modeled by considering transport behaviors oxygen vacancies ions together with temperature effect. Both metallic-like electron hopping conduction transports are considered RRAM. The can reproduce both typical I-V characteristics RRAM in...

Neuromorphic computing is an attractive computation paradigm that complements the von Neumann architecture. The salient features of neuromorphic are massive parallelism, adaptivity to complex input information, and tolerance errors. As one most crucial components in a system, electronic synapse requires high device integration density low-energy consumption. Oxide-based resistive switching devices have been shown be promising candidate realize functions synapse. However, intrinsic variation...

The resistive random access memory (RRAM) crossbar array has been extensively studied as one of the most promising candidates for future high-density nonvolatile technology. However, some problems caused by circuit and device interaction, such sneak leakage paths, result in limited size large power consumption, which degrade performance significantly. Thus, analysis on interaction issue is imperative. In this paper, a simulation method developed to investigate critical issues correlated with...

Resistive switching processes in HfO2 are studied by electron holography and situ energy-filtered imaging. The results show that oxygen vacancies gradually generated the oxide layer under ramped electrical bias, finally form several conductive channels connecting two electrodes. It also shows process occurs at top interface of hafnia layer. As a service to our authors readers, this journal provides supporting information supplied authors. Such materials peer reviewed may be re-organized for...

We report, for the first time, three types of endurance failure behaviors in TMO based RRAM. New physical mechanisms are proposed to clarify origins these failures. A physically-based optimized switching mode is developed improve TMO-RRAM. significantly enhanced >;10 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">9</sup> cycles was demonstrated HfO <sub xmlns:xlink="http://www.w3.org/1999/xlink">x</sub> /TiO /HfO devices.

Double-layer stacked HfOx vertical RRAM is demonstrated for 3D cross-point architecture using a cost-effective fabrication process. Electrode/oxide interface engineering TiON layer results in non-linear I-V suitable the selector-less array. The fabricated shows excellent performances such as reset current (<;50μA), switching speed (~50ns), endurance (>10 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">8</sup> cycles), half-selected read...

A SPICE model of oxide-based resistive random access memory (RRAM) for dc and transient behaviors is developed based on the conductive filament evolution implemented in large-scale array simulation. The simulations one transistor-one resistor RRAM up to 16 kb with wire resistance (R <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">wire</sub> ) capacitance (C indicate that: 1) resistance-capacitance delay during RESET leakage current SET have...

Resistive random-access memory devices with atomic layer deposition HfO2 and radio frequency sputtering TiOx as resistive switching layers were fabricated successfully. Low-power characteristic 1.52 μW set power (1 μA@1.52 V) 1.12 reset μA@1.12 was obtained in the HfO2/TiOx (RRAM) by controlling oxygen content of layer. Besides, influence during process on properties would be discussed detail. The investigations indicated that "soft breakdown" occurred easily electrical forming/set RRAM high...

Abstract Developing energy-efficient parallel information processing systems beyond von Neumann architecture is a long-standing goal of modern technologies. The widely used computer separates memory and computing units, which leads to energy-hungry data movement when computers work. In order meet the need efficient for data-driven applications such as big Internet Things, an critical society. Here we show non-von built resistive switching (RS) devices named “iMemComp”, where logic are...

Resistive switching random access memory (RRAM) is a leading candidate for next-generation nonvolatile and storage-class memories monolithic integration of logic with interleaved in multiple layers. To meet the increasing need device-circuit-system co-design optimization applications from digital systems to brain-inspired computing systems, SPICE model RRAM that can reproduce essential device physics circuit simulation environment required. In this work, we develop an capture all...

Abstract The finding of the robust ferroelectricity in HfO 2 -based thin films is fantastic from view point both fundamentals and applications. In this review article, current research status future prospects for ferroelectric devices are presented to related issues discussed, which include: 1) characteristics observed associated with factors dopant, strain, interface, thickness, defect, fabrication condition, more; 2) physical understanding on behaviors by density functional theory...

A unified microscopic principle is proposed to clarify resistive switching behaviors of transition metal oxide based random access memories (RRAM) for the first time. In this principle, both unipolar and bipolar characteristics RRAM are correlated with distribution localized oxygen vacancies in layer, which governed by generation recombination dissociative ions. Based on an atomistic simulation method developed evaluate critical memory performance, successfully conduct device optimization....

An analytic model for the endurance degradation of metal oxide based RRAM is presented first time. The behaviors under various operation modes can be predicted by model, which were verified measured data in different devices. Furthermore, 10 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">6</sup> all 4-level resistance states HfO <sub xmlns:xlink="http://www.w3.org/1999/xlink">X</sub> -based demonstrated using proposed optimization scheme...

Although bi-directional analog switching capability is crucial for neuromorphic computing application, it still difficult to be realized in filamentary RRAM cells. This work investigates the physical mechanism of abrupt transition using Kinetic Monte Carlo simulation method. A disorder-related model oxygen vacancy distribution proposed with an order parameter Oy quantify behaviors different devices. The results and predictions are verified by experiments performed on 1kb array. It suggested...

An analog neuromorphic system is developed based on the fabricated resistive switching memory array. A novel training scheme proposed to optimize performance of by utilizing segmented synaptic behavior. The demonstrated a grayscale image recognition. According experiment results, optimized one improves learning accuracy from 77.83% 91.32%, decreases energy consumption more than two orders, and substantially boosts efficiency compared traditional scheme.

HfOx-based resistive switching device has been explored as one of the promising candidates for electronic synapses neuromorphic computing systems due to its high performance, low cost, and compatibility with CMOS technology. To meet codesign requirement neurons in systems, a compact model that can capture synaptic futures is developed. The developed accurately describe multilevel conductance transition behaviors during RESET process depression learning well binary stochastic behavior SET...

In this paper, a hardware-realized neuromorphic system for pattern recognition is presented. The directly captures images from the environment, and then conducts classification using single layer neural network. Metal-oxide resistive random access memory (RRAM) used as electronic synapses, threshold-controlled neurons are proposed postsynaptic to save area simplify operation. neuron, no capacitor utilized, which contributes higher integration density. total energy consumption of RRAM...

Nonvolatile and cascadable stateful logic operations are experimentally demonstrated within a 1 k-bit one-transistor-one-resistor (1T1R) resistive random access memory (RRAM) array, where NAND gates serve as the building blocks. A robust dual-gate-voltage operation scheme is proposed. The effects of transistor ON robustness to device parameter variations discussed. parallel 4-bit bitwise XOR implemented in 1T1R array by cascading gates. This letter presents feasible approach in-memory...

In this work, an evaluation methodology is proposed to study the impacts of state instability and retention failure filamentary analog resistive random access memory (FA-RRAM) on performance deep neural networks (DNNs). Based methodology, analytic model for statistical behaviors applied evaluate reliability FA-RRAM 11-layer FA-RRAM-based DNN CIFAR-10 recognition. Simulations indicate that recognition accuracy decreases rapidly with increase baking time (t = 10 <sup...

In this work, a new operation scheme is developed to improve the state stability of multi-level resistive random-access memory (RRAM) array. We found that instability after programming mainly derived from excessive oxygen vacancies generated by abrupt SET process. Based on understanding operation, triangular pulse adopted for RRAM, which can effectively suppress short-term relaxation and long-time retention. Comparison results show degradation recognition accuracy Cifar-10 over time in...