RESUMEN

A nonlinear system, exhibiting a unique asymptotic behaviour, while being continuously subject to a stimulus from a certain class, is said to suffer from fading memory. This interesting phenomenon was first uncovered in a non-volatile tantalum oxide-based memristor from Hewlett Packard Labs back in 2016 out of a deep numerical investigation of a predictive mathematical description, known as the Strachan model, later corroborated by experimental validation. It was then found out that fading memory is ubiquitous in non-volatile resistance switching memories. A nonlinear system may however also exhibit a local form of fading memory, in case, under an excitation from a given family, it may approach one of a number of distinct attractors, depending upon the initial condition. A recent bifurcation study of the Strachan model revealed how, under specific train stimuli, composed of two square pulses of opposite polarity per cycle, the simplest form of local fading memory affects the transient dynamics of the aforementioned Resistive Random Access Memory cell, which, would asymptotically act as a bistable oscillator. In this manuscript we propose an analytical methodology, based on the application of analysis tools from Nonlinear System Theory to the Strachan model, to craft the properties of a generalised pulse train stimulus in such a way to induce the emergence of complex local fading memory effects in the nano-device, which would consequently display an interesting tuneable multistable oscillatory response, around desired resistance states. The last part of the manuscript discusses a case study, shedding light on a potential application of the local history erase effects, induced in the device via pulse train stimulation, for compensating the unwanted yet unavoidable drifts in its resistance state under power off conditions.

RESUMEN

We studied the phase change and resistive switching characteristics of copper oxide (CuxO) films through post-thermal annealing. This investigation aimed to assess the material's potential for a variety of electrical devices, exploring its versatility in electronic applications. The CuxO films deposited by RF magnetron sputtering were annealed at 300, 500, and 700 °C in ambient air for 4 min by rapid thermal annealing (RTA) method, and then it was confirmed that the structural phase change from Cu2O to CuO occurred with increasing annealing temperature. Resistive random-access memory (ReRAM) devices with Au/CuxO/p+-Si structures were fabricated, and the ReRAM properties appeared in CuO-based devices, while Cu2O ReRAM devices did not exhibit resistive switching behavior. The CuO ReRAM device annealed at 500 °C showed the best properties, with a on/off ratio of 8 × 102, good switching endurance of ∼100 cycles, data retention for 104s, and stable uniformity in the cumulative probability distribution. This characteristic change could be explained by the difference in the grain size and density of defects between the Cu2O and CuO films. These results demonstrate that superior and stable resistive switching properties of RF-sputtered CuxO films can be obtained by low-temperature RTA.

RESUMEN

In this study, resistive random-access memory (ReRAM) devices with ZnO nanoparticles (NPs) are suggested to enhance performance and reduce variation in device switching parameters. The ZnO NPs are formed by annealing ZnO prepared via atomic layer deposition on HfO2, which is verified using transmission electron microscopy, x-ray diffraction pattern, and atomic force microscopy. The depth profile analysis of x-ray photoelectron spectroscopy shows that oxygen diffuses from HfO2to ZnO NPs during annealing. This can be explained by the calculation results using density functional theory (DFT) where the formation energy of oxygen vacancies is reduced at the interface of ZnO NPs and HfO2compared to single HfO2. The fabricated ZnO NPs ReRAM demonstrates reduced forming voltage, stable resistive switching behavior, and improved cycle-to-cycle uniformity in a high-resistance state.

Asunto(s)

RESUMEN

The performance stability of the resistive switching (RS) is vital for a resistive random-access memory device. Here, by inserting a thin HfAlOxlayer between the InGaZnO (IGZO) layer and the bottom Pt electrode, the RS performance in amorphous IGZO memory device is significantly improved. Comparing with a typical metal-insulator-metal structure, the device with HfAlOxlayer exhibits lower switching voltages, faster switching speeds, lower switching energy and lower power consumption. As well, the uniformity of switching voltage and resistance state is also improved. Furthermore, the device with HfAlOxlayer exhibits long retention time (>104s at 85 °C) , high on/off ratio and more than 103cycles of endurance at atmospheric environment. Those substantial improvements in IGZO memory device are attributed to the interface effects with a HfAlOxinsertion layer. With such layer, the formation and rupture locations of Ag conductive filaments are better regulated and confined, thus an improved performance stability.

RESUMEN

This paper proposes two different approaches to studying resistive switching of oxide thin films using scratching probe nanolithography of atomic force microscopy (AFM). These approaches allow us to assess the effects of memristor size and top-contact thickness on resistive switching. For that purpose, we investigated scratching probe nanolithography regimes using the Taguchi method, which is known as a reliable method for improving the reliability of the result. The AFM parameters, including normal load, scratch distance, probe speed, and probe direction, are optimized on the photoresist thin film by the Taguchi method. As a result, the pinholes with diameter ranged from 25.4 ± 2.2 nm to 85.1 ± 6.3 nm, and the groove array with a depth of 40.5 ± 3.7 nm and a roughness at the bottom of less than a few nanometers was formed. Then, based on the Si/TiN/ZnO/photoresist structures, we fabricated and investigated memristors with different spot sizes and TiN top contact thickness. As a result, the HRS/LRS ratio, USET, and ILRS are well controlled for a memristor size from 27 nm to 83 nm and ranged from ~8 to ~128, from 1.4 ± 0.1 V to 1.8 ± 0.2 V, and from (1.7 ± 0.2) × 10-10 A to (4.2 ± 0.6) × 10-9 A, respectively. Furthermore, the HRS/LRS ratio and USET are well controlled at a TiN top contact thickness from 8.3 ± 1.1 nm to 32.4 ± 4.2 nm and ranged from ~22 to ~188 and from 1.15 ± 0.05 V to 1.62 ± 0.06 V, respectively. The results can be used in the engineering and manufacturing of memristive structures for neuromorphic applications of brain-inspired artificial intelligence systems.

RESUMEN

A new architecture has become necessary owing to the power consumption and latency problems of the von Neumann architecture. A neuromorphic memory system is a promising candidate for the new system as it has the potential to process large amounts of digital information. A crossbar array (CA), which consists of a selector and a resistor, is the basic building block for the new system. Despite the excellent prospects of crossbar arrays, the biggest obstacle for them is sneak current, which can cause a misreading between the adjacent memory cells, thus resulting in a misoperation in the arrays. The chalcogenide-based ovonic threshold switch (OTS) is a powerful selector with highly nonlinear I-V characteristics that can be used to address the sneak current problem. In this study, we evaluated the electrical characteristics of an OTS with a TiN/GeTe/TiN structure. This device shows nonlinear DC I-V characteristics, an excellent endurance of up to 109 in the burst read measurement, and a stable threshold voltage below 15 mV/dec. In addition, at temperatures below 300 °C, the device exhibits good thermal stability and retains an amorphous structure, which is a strong indication of the aforementioned electrical characteristics.

RESUMEN

By imitating the synaptic connectivity and plasticity of the brain, emerging electronic nanodevices offer new opportunities as the building blocks of neuromorphic systems. One challenge for large-scale simulations of computational architectures based on emerging devices is to accurately capture device response, hysteresis, noise, and the covariance structure in the temporal domain as well as between the different device parameters. We address this challenge with a high throughput generative model for synaptic arrays that is based on a recently available type of electrical measurement data for resistive memory cells. We map this real-world data onto a vector autoregressive stochastic process to accurately reproduce the device parameters and their cross-correlation structure. While closely matching the measured data, our model is still very fast; we provide parallelized implementations for both CPUs and GPUs and demonstrate array sizes above one billion cells and throughputs exceeding one hundred million weight updates per second, above the pixel rate of a 30 frames/s 4K video stream.

RESUMEN

In organic resistive random-access memory (ReRAM) devices, deeply understanding how to control the performance of π-conjugated semiconductors through molecular-shape-engineering is important and highly desirable. Herein, we design a family of N-containing heteroaromatic semiconductors with molecular shapes moving from mono-branched 1Q to di-branched 2Q and tri-branched 3Q. We find that this molecular-shape engineering can induce reliable binary to ternary ReRAM switching, affording a highly enhanced device yield that satisfies the practical requirement. The density functional theory calculation and experimental evidence suggest that the increased multiple paired electroactive nitrogen sites from mono-branched 1Q to tri-branched 3Q are responsible for the multilevel resistance switching, offering stable bidentate coordination with the active metal atoms. This study sheds light on the prospect of N-containing heteroaromatic semiconductors for promising ultrahigh-density data-storage ReRAM application.

RESUMEN

Processing-in-memory (PIM) is emerging as a new computing paradigm to replace the existing von Neumann computer architecture for data-intensive processing. For the higher end-user mobility, low-power operation capability is more increasingly required and components need to be renovated to make a way out of the conventional software-driven artificial intelligence. In this work, we investigate the hardware performances of PIM architecture that can be presumably constructed by resistive-switching random-access memory (ReRAM) synapse fabricated with a relatively larger thermal budget in the full Si processing compatibility. By introducing a medium-temperature oxidation in which the sputtered Ge atoms are oxidized at a relatively higher temperature compared with the ReRAM devices fabricated by physical vapor deposition at room temperature, higher device reliability has been acquired. Based on the empirically obtained device parameters, a PIM architecture has been conceived and a system-level evaluations have been performed in this work. Considerations include the cycle-to-cycle variation in the GeOx ReRAM synapse, analog-to-digital converter resolution, synaptic array size, and interconnect latency for the system-level evaluation with the Canadian Institute for Advance Research-10 dataset. A fully Si processing-compatible and robust ReRAM synapse and its applicability for PIM are demonstrated.

RESUMEN

With the resistive random access memory (ReRAM) devices based on the Al/BaTiO3 (BTO)/ITO structure fabricated at hand, by cross-analyzing the resistive memory characteristics in terms of various barium titanate (BTO) film thicknesses, it is found that the device with 60 nm thick BTO can be switched more than 425 times, while the corresponding SET/RESET voltage, the on-off ratio, and the retention time are -0.69 V/0.475 V, 102, and more than 104 seconds, respectively. Furthermore, the aforementioned ReRAM with a low switching voltage and low power consumption is further integrated with a waveguide resonator in the form of a dual microdisk aligned in a parallel fashion. As the separation gap between the two microdisks is fixed at 15 µm, the ReRAM-mediated dual disk resonator would render a 180° phase reversal between the spectral outputs of the through-port and drop-port. If the gap is shortened to 10 and 5 µm, the expected phase reversal could also be retrieved due to the selective combinations of different memory states associated with each of the two ReRAM microdisks as witnessed by a series of characterization measurements.

RESUMEN

Emerging resistive random-access memory (ReRAM) has demonstrated great potential in the achievement of the in-memory computing paradigm to overcome the well-known "memory wall" in current von Neumann architecture. The ReRAM crossbar array (RCA) is a promising circuit structure to accelerate the vital multiplication-and-accumulation (MAC) operations in deep neural networks (DNN). However, due to the nonlinear distribution of conductance levels in ReRAM, a large deviation exists in the mapping process when the trained weights that are quantized by linear relationships are directly mapped to the nonlinear conductance values from the realistic ReRAM device. This deviation degrades the inference accuracy of the RCA-based DNN. In this paper, we propose a minimum error substitution based on a conductance-aware quantization method to eliminate the deviation in the mapping process from the weights to the actual conductance values. The method is suitable for multiple ReRAM devices with different non-linear conductance distribution and is also immune to the device variation. The simulation results on LeNet5, AlexNet and VGG16 demonstrate that this method can vastly rescue the accuracy degradation from the non-linear resistance distribution of ReRAM devices compared to the linear quantization method.

RESUMEN

In recent years, compute-in-memory (CIM) has been extensively studied to improve the energy efficiency of computing by reducing data movement. At present, CIM is frequently used in data-intensive computing. Data-intensive computing applications, such as all kinds of neural networks (NNs) in machine learning (ML), are regarded as 'soft' computing tasks. The 'soft' computing tasks are computations that can tolerate low computing precision with little accuracy degradation. However, 'hard' tasks aimed at numerical computations require high-precision computing and are also accompanied by energy efficiency problems. Numerical computations exist in lots of applications, including partial differential equations (PDEs) and large-scale matrix multiplication. Therefore, it is necessary to study CIM for numerical computations. This article reviews the recent developments of CIM for numerical computations. The different kinds of numerical methods solving partial differential equations and the transformation of matrixes are deduced in detail. This paper also discusses the iterative computation of a large-scale matrix, which tremendously affects the efficiency of numerical computations. The working procedure of the ReRAM-based partial differential equation solver is emphatically introduced. Moreover, other PDEs solvers, and other research about CIM for numerical computations, are also summarized. Finally, prospects and the future of CIM for numerical computations with high accuracy are discussed.

RESUMEN

TiN/AlOx:Ti/TaOx/TiN memory devices using bilayer resistive switching memory demonstrated excellent durability and capability of QLC (quad-level cell) memory devices. The best nonvolatile memory characteristics with the lowest operation current and optimized 4 bit/cell states were obtained using the Incremental Step Pulse Programming (ISPP) algorithm in array. As a result, a superior QLC reliability (cycle endurance > 1 k at each level of the QLC, data retention > 2 h at 125 °C) for all the 4 bits/cell operations was achieved in sub-µm scaled RRAM (resistive random access memory) devices.

RESUMEN

Resistive switching induced by ion migration is promising for applications such as random-access memory (ReRAM) and neuromorphic transistors. Hydride ions (H-) are an interesting candidate as the migration ion for resistive switching devices because they have fast diffusion in several compounds at room temperature and doping/dedoping can be used effectively to achieve significant changes in the electronic conductivity. Here, we report reversible resistive switching characteristics in rare-earth oxyhydrides (REHxO(3-x)/2) induced by field insertion/extraction of H-. The current-voltage measurements revealed that the resistive switching response, hysteresis, and switching voltage vary greatly with the H-/O2- ratio in the films. We fabricated a ReRAM device using Ti/YH1.3O0.85/MoOx structure and confirmed the bipolar-type operation with the resistance switching ratio of 1 order of magnitude over 1000 cycles. The composition gradient of H-/O2- in YHxO(3-x)/2 films, in addition to the hydrogen-absorbing ability of the top electrode, is essential for effective device operation. Our findings show that hydride-conducting solid-state electrolytes are suitable for resistive switching device development.

RESUMEN

This article presents the results of experimental studies of the impact of electrode material and the effect of nanoscale film thickness on the resistive switching in forming-free nanocrystalline ZnO films grown by pulsed laser deposition. It was demonstrated that the nanocrystalline ZnO film with TiN, Pt, ZnO:In, and ZnO:Pd bottom electrodes exhibits a nonlinear bipolar effect of forming-free resistive switching. The sample with Pt showed the highest resistance values RHRS and RLRS and the highest value of Uset = 2.7 ± 0.4 V. The samples with the ZnO:In and ZnO:Pd bottom electrode showed the lowest Uset and Ures values. An increase in the number of laser pulses from 1000 to 5000 was shown to lead to an increase in the thickness of the nanocrystalline ZnO film from 7.2 ± 2.5 nm to 53.6 ± 18.3 nm. The dependence of electrophysical parameters (electron concentration, electron mobility, and resistivity) on the thickness of the forming-free nanocrystalline ZnO film for the TiN/ZnO/W structure was investigated. The endurance test and homogeneity test for TiN/ZnO/W structures were performed. The structure Al2O3/TiN/ZnO/W with a nanocrystalline ZnO thickness 41.2 ± 9.7 nm was shown to be preferable for the manufacture of ReRAM and memristive neuromorphic systems due to the highest value of RHRS/RLRS = 2307.8 ± 166.4 and low values of Uset = 1.9 ± 0.2 V and Ures = -1.3 ± 0.5 V. It was demonstrated that the use of the TiN top electrode in the Al2O3/TiN/ZnO memristor structure allowed for the reduction in Uset and Ures and the increase in the RHRS/RLRS ratio. The results obtained can be used in the manufacturing of resistive-switching nanoscale devices for neuromorphic computing based on the forming-free nanocrystalline ZnO oxide films.

RESUMEN

While promising for high-capacity machine learning accelerators, memristor devices have non-idealities that prevent software-equivalent accuracies when used for online training. This work uses a combination of Mini-Batch Gradient Descent (MBGD) to average gradients, stochastic rounding to avoid vanishing weight updates, and decomposition methods to keep the memory overhead low during mini-batch training. Since the weight update has to be transferred to the memristor matrices efficiently, we also investigate the impact of reconstructing the gradient matrixes both internally (rank-seq) and externally (rank-sum) to the memristor array. Our results show that streaming batch principal component analysis (streaming batch PCA) and non-negative matrix factorization (NMF) decomposition algorithms can achieve near MBGD accuracy in a memristor-based multi-layer perceptron trained on the MNIST (Modified National Institute of Standards and Technology) database with only 3 to 10 ranks at significant memory savings. Moreover, NMF rank-seq outperforms streaming batch PCA rank-seq at low-ranks making it more suitable for hardware implementation in future memristor-based accelerators.

RESUMEN

Conductive filaments (CFs) play a critical role in the mechanism of resistive random-access memory (ReRAM) devices. However, in situ detection and visualization of the precise location of CFs are still key challenges. We demonstrate for the first time the use of a π-conjugated molecule which can transform between its twisted and planar states upon localized Joule heating generated within filament regions, thus reflecting the locations of the underlying CFs. Customized patterns of CFs were induced and observed by the π-conjugated molecule layer, which confirmed the hypothesis. Additionally, statistical studies on filaments distribution were conducted to study the effect of device sizes and bottom electrode heights, which serves to enhance the understanding of switching behavior and their variability at device level. Therefore, this approach has great potential in aiding the development of ReRAM technology.

RESUMEN

Major challenges concerning the reliability of resistive switching random access memories based on the valence change mechanism (VCM) are short-term instability and long-term retention failure of the programmed resistance state, particularly in the high resistive state. On the one hand, read noise limits the reliability of VCMs via comparatively small current jumps especially when looking at the statistics of millions of cells that are needed for industrial applications. Additionally, shaping algorithms aiming for an enlargement of the read window are observed to have no lasting effect. On the other hand, long-term retention failures limiting the lifetime of the programmed resistance states need to be overcome. The physical origin of these phenomena is still under debate and needs to be understood much better. In this work, we present a three-dimensional kinetic Monte Carlo simulation model where we implemented diffusion-limiting domains to the oxide layer of the VCM cell. We demonstrate that our model can explain both instability and retention failure consistently by the same physical processes. Further, we find that the random diffusion of oxygen vacancies plays an important role regarding the reliability of VCMs and can explain instability phenomena as the shaping failure as well as the long-term retention failure in our model. Additionally, the results of the simulations are compared with experimental data of read noise and retention investigations on ZrO2-based VCM devices.

RESUMEN

With the arrival of the Internet of Things (IoT) and the challenges arising from Big Data, neuromorphic chip concepts are seen as key solutions for coping with the massive amount of unstructured data streams by moving the computation closer to the sensors, the so-called "edge computing." Augmenting these chips with emerging memory technologies enables these edge devices with non-volatile and adaptive properties which are desirable for low power and online learning operations. However, an energy- and area-efficient realization of these systems requires disruptive hardware changes. Memristor-based solutions for these concepts are in the focus of research and industry due to their low-power and high-density online learning potential. Specifically, the filamentary-type valence change mechanism (VCM memories) have shown to be a promising candidate In consequence, physical models capturing a broad spectrum of experimentally observed features such as the pronounced cycle-to-cycle (c2c) and device-to-device (d2d) variability are required for accurate evaluation of the proposed concepts. In this study, we present an in-depth experimental analysis of d2d and c2c variability of filamentary-type bipolar switching HfO2/TiOx nano-sized crossbar devices and match the experimentally observed variabilities to our physically motivated JART VCM compact model. Based on this approach, we evaluate the concept of parallel operation of devices as a synapse both experimentally and theoretically. These parallel synapses form a synaptic array which is at the core of neuromorphic chips. We exploit the c2c variability of these devices for stochastic online learning which has shown to increase the effective bit precision of the devices. Finally, we demonstrate that stochastic switching features for a pattern classification task that can be employed in an online learning neural network.

RESUMEN

Halide perovskite materials such as methylammonium lead iodide (CH3NH3PbI3) have attracted considerable interest for the resistive random-access memory applications, which exploit a dramatic change in the resistance by an external electric bias. In many semiconductor films, the drift, accumulation, and chain formation of defects explain the change in the resistance by an external bias. This study demonstrates that the interface of CH3NH3PbI3 with TiO2 has a significant impact on the formation and rupture of defect chains and causes the asymmetric bipolar resistive switching in the Au/CH3NH3PbI3/TiO2/FTO device (FTO = fluorine-doped tin oxide). When a negative bias is applied to the Au electrode, iodine interstitials with the lowest migration activation energy move toward TiO2 in the CH3NH3PbI3 layer and pile up at the CH3NH3PbI3-TiO2 interface. Under the same condition, oxygen vacancies in the TiO2 layer also travel to the CH3NH3PbI3-TiO2 interface and strongly attract iodine interstitials. As a result, a Schottky barrier appears at the CH3NH3PbI3-TiO2 interface, and the resistance of Au/CH3NH3PbI3/TiO2/FTO becomes much larger than that of Au/CH3NH3PbI3/FTO in the high resistance state. The frequency dependence of the capacitance confirms the asymmetric appearance of a large space charge polarization at the CH3NH3PbI3-TiO2 interface, which causes the unique bipolar resistive switching behavior with the on/off ratio (103) and retention time (>104 seconds) at -0.85 V in Au/CH3NH3PbI3/TiO2/FTO film.