FG Halbleitermaterialien
Refine
Year of publication
Document Type
Way of publication
- Open Access (11)
Language
- English (142)
Keywords
- RRAM (42)
- neural network (13)
- HfO2 (12)
- Multilevel switching (11)
- Graphene (9)
- memristive device (9)
- memristive switching (4)
- resistive switching (4)
- Neural network (3)
- Transistor (3)
Institute
Towards the Growth of Hexagonal Boron Nitride on Ge(001)/Si Substrates by Chemical Vapor Deposition
(2022)
The growth of hexagonal boron nitride (hBN) on epitaxial Ge(001)/Si substrates via high-vacuum chemical vapor deposition from borazine is investigated for the first time in a systematic manner. The influences of the process pressure and growth temperature in the range of 10−7–10−3 mbar and 900–980 °C, respectively, are evaluated with respect to morphology, growth rate, and crystalline quality of the hBN films. At 900 °C, nanocrystalline hBN films with a lateral crystallite size of ~2–3 nm are obtained and confirmed by high-resolution transmission electron microscopy images. X-ray photoelectron spectroscopy confirms an atomic N:B ratio of 1 ± 0.1. A three-dimensional growth mode is observed by atomic force microscopy. Increasing the process pressure in the reactor mainly affects the growth rate, with only slight effects on crystalline quality and none on the principle growth mode. Growth of hBN at 980 °C increases the average crystallite size and leads to the formation of 3–10 well-oriented, vertically stacked layers of hBN on the Ge surface. Exploratory ab initio density functional theory simulations indicate that hBN edges are saturated by hydrogen, and it is proposed that partial de-saturation by H radicals produced on hot parts of the set-up is responsible for the growth
The crossbar structure of Resistive-switching random access memory (RRAM) arrays enabled the In-Memory Computing circuits paradigm, since they imply the native acceleration of a crucial operations in this scenario, namely the Matrix-Vector-Multiplication (MVM). However, RRAM arrays are affected by several issues materializing in conductance variations that might cause severe performance degradation. A critical
one is related to the drift of the low conductance states appearing
immediately at the end of program and verify algorithms that
are mandatory for an accurate multi-level conductance operation. In this work, we analyze the benefits of a new programming
algorithm that embodies Set and Reset switching operations to
achieve better conductance control and lower variability. Data
retention analysis performed with different temperatures for 168
hours evidence its superior performance with respect to standard programming approach. Finally, we explored the benefits
of using our methodology at a higher abstraction level, through
the simulation of an Artificial Neural Network for image recognition task (MNIST dataset). The accuracy achieved shows higher
performance stability over temperature and time.
A graphene-based three terminal barristor device was proposed to overcome the low on/off ratios and insufficient current saturation of conventional graphene field effect transistors. In this study, we fabricated and analyzed a novel graphene-based transistor, which resembles the structure of the barristor but uses a different operating condition. This new device, termed graphene adjustable-barriers transistor (GABT), utilizes a semiconductor-based gate rather than a metal−insulator gate structure to modulate the device currents. The key feature of the device is the two graphene-semiconductor Schottky barriers with different heights that are controlled simultaneously by the gate voltage. Due to the asymmetry of the barriers, the drain current exceeds the gate current by several orders of magnitude. Thus, the GABT can be considered an amplifier with an alterable current gain. In this work, a silicon−graphene−germanium GABT with an ultra-high current gain (ID/IG up to 8 × 106) was fabricated, and the device functionality was demonstrated. Additionally, a capacitance model is applied to predict the theoretical device performance resulting in an on−off ratio above 106, a swing of 87 mV/dec, and a drivecurrent of about 1 × 106 A/cm2.
One of the objectives fostered in medical science is the so-called precision medicine, which requires the analysis of a large amount of survival data from patients to deeply understand treatment options. Tools like Machine Learning and Deep Neural Networks are becoming a de-facto standard. Nowadays, computing facilities based on the Von Neumann architecture are devoted to these tasks, yet rapidly hitting a bottleneck in performance and energy efficiency. The In-Memory Computing (IMC) architecture emerged as a revolutionary approach to overcome that issue. In this work, we propose an IMC architecture based on Resistive switching memory (RRAM) crossbar arrays to provide a convenient primitive for matrix–vector multiplication in a single computational step. This opens massive performance improvement in the acceleration of a neural network that is frequently used in survival analysis of biomedical records, namely the DeepSurv. We explored how the synaptic weights mapping strategy and the programming algorithms developed to counter RRAM non-idealities expose a performance/energy trade-off. Finally, we assessed the benefits of the proposed architectures with respect to a GPU-based realization of the same task, evidencing a tenfold improvement in terms of performance and three orders of magnitude with respect to energy efficiency.
The impact of artificial intelligence on human life
has increased significantly in recent years. However, as the
complexity of problems rose aswell, increasing system features
for such amount of data computation became troublesome due
to the von Neumann’s computer architecture. Neuromorphic
computing aims to solve this problem by mimicking the parallel
computation of a human brain. For this approach, memristive
devices are used to emulate the synapses of a human brain. Yet,
common simulations of hardware based networks require time
consuming Monte-Carlo simulations to take into account the
stochastic switching of memristive devices. This work presents
an alternative concept making use of the convolution of the
probability distribution functions (PDF) of memristor currents
by its equivalent multiplication in Fourier domain. An artificial
neural network is accordingly implemented to perform the
inference stage with handwritten digits.
Due to the unique electronic structures, graphene and other 2D Materials are considered as materials which can enable and extend the functionalities and performance in a large variety of applications, among them in microelectronics. At this point, the investigation and preparation of graphene devices in conditions resembling as close as possible the Si technology environment is of highest importance.Towards these goals, this paper focuses on the full spectra of graphene research aspects in 200mm pilot line. We investigated different process module developments such as CMOS compatible growth of high quality graphene on germanium and its growth mechanisms, transfer related challenges on target substrates, patterning, passivation and various concepts of contacting of graphene on a full 200 mm wafers. Finally, we fabricated proof-of-concept test structures e.g. TLM, Hall bars and capacitor structures to prove the feasibility of graphene processing in the pilot line of IHP.
Accomplishing truly analog conductance modulation in memristive arrays is crucial in order to implement the synaptic plasticity in hardware-based neuromorphic systems. In this paper, such a feature was addressed by exploiting the inherent stochasticity of switching dynamics in amorphous HfO2 technology. A thorough statistical analysis of experimental characteristics measured in 4 kbit arrays by using trains of identical depression/potentiation pulses with different voltage amplitudes and pulse widths provided the key to develop two different updating rules and to define their optimal programming parameters. The first rule is based on applying a specific number of identical pulses until the conductance value achieves the
desired level. The second one utilized only one single pulse with a particular amplitude to achieve the targeted conductance level. In addition, all the results provided by the statistical analysis performed may play an important role in understanding better the switching behavior of this particular technology.
Mitigating the Effects of RRAM Process Variation on the Accuracy of Artifical Neural Networks
(2022)
Weight storage is a key challenge in the efficient implementation of artificial neural networks. Novel memory technologies such as RRAM are able to greatly improve density and introduce non-volatility and multibit capabilities to this component of ANN accelerators. The usage of RRAM in this domain comes with downsides, mainly caused by cycle-to-cycle and device-to-device variability leading to erroneous readouts, greatly affecting digital systems. ANNs have the ability to compensate for this by their inherent redundancy and usually exhibit a gradual deterioration in the accuracy of the task at hand. This means, that slight error rates can be acceptable for weight storage in an ANN accelerator. In this work we link device-to-device variability to the accuracy of an ANN for such an accelerator. From this study, we can estimate how strongly a certain net is affected by a certain device parameter variability. This methodology is then used to present three mitigation strategies and to evaluate how they affect the reaction of the network to variability: a) Dropout Layers b) Fault-Aware Training c) Redundancy. These mitigations are then evaluated by their ability to improve accuracy and to lower hardware overhead by providing data for a real-word example. We improved this network’s resilience in such a way that it could tolerate double the variation in one of the device parameters (standard deviation of the oxide thickness can be 0.4 nm instead of 0.2 nm while maintaining sufficient accuracy.)
Resistive-switching random access memory (RRAM) is a promising technology for in-memory computing (IMC) to accelerate training and inference of deep neural networks (DNNs). This work presents the first physics-based statistical model describing (i) multilevel RRAM device program/verify (PV) algorithms by controlled set transition, (ii) the stochastic cycle-to-cycle (C2C) and device-to-device (D2D) variations within the array, and (iii) the impact of such imprecisions on the accuracy of DNN accelerators. The model can handle the full chain from RRAM materials/device parameters to the DNN performance, thus providing a valuable tool for device/circuit codesign of hardware DNN accelerators.
The resistive switching properties of HfO2 based 1T-1R memristive devices are electrically modified by adding ultra-thin layers of Al2O3
into the memristive device. Three different types of memristive stacks are fabricated in the 130 nm CMOS technology of IHP. The switching properties of the memristive devices are discussed with respect to forming voltages, low resistance state and high resistance state characteristics and their variabilities. The experimental I–V characteristics of set and reset operations are evaluated by using the quantum point contact model. The properties of the conduction filament in the on and off states of the memristive devices are discussed with respect to the model parameters obtained from the QPC fit.
The use of alternating current (AC) electrokinetic forces, like dielectrophoresis and AC electroosmosis, as a simple and fast method to immobilize sub-micrometer objects onto nanoelectrode arrays is presented. Due to its medical relevance, the influenza virus is chosen as a model organism. One of the outstanding features is that the immobilization of viral material to the electrodes can be achieved permanently, allowing subsequent handling independently from the electrical setup. Thus, by using merely electric fields, we demonstrate that the need of prior chemical surface modification could become obsolete. The accumulation of viral material over time is observed by fluorescence microscopy. The influences of side effects like electrothermal fluid flow, causing a fluid motion above the electrodes and causing an intensity gradient within the electrode array, are discussed. Due to the improved resolution by combining fluorescence microscopy with deconvolution, it is shown that the viral material is mainly drawn to the electrode edge and to a lesser extent to the electrode surface. Finally, areas of application for this functionalization technique are presented.
We present a statistical procedure for the extraction of parameters of a compact model for memristive devices. Thereby, in a circuit simulation the typical fluctuations of the current–voltage (I-V) characteristics from device-to-device (D2D) and from cycle-to-cycle (C2C) can be emulated. The approach is based on the Stanford model
whose parameters play a key role to integrating D2D and C2C dispersion. The influence of such variabilities over the model’s parameters is investigated by using a fitting algorithm fed with experimental data. After this, the statistical distributions of the parameters are used in a Monte Carlo simulation to reproduce the I-V D2D and C2C dispersions which show a good agreement to the measured curves. The results allow the simulation of the on/off current variation for the design of RRAM cells or memristor-based artificial neural networks.
A new kind of transistor device with a graphene monolayer embedded between two n-type silicon layers is fabricated and characterized. The device is called graphene-base heterojunction transistor (GBHT). The base-voltage controls the current of the device flowing from the emitter via graphene to the collector. The transit time for electrons passing by the ultrathin graphene layer is extremely short which makes the device very promising for high frequency RF-electronics. The output current of the device is saturated and clearly modulated by the base voltage. Further, the silicon collector of the GBHT is replaced by germanium to improve the device performance. This enabled the collector current to be increased by almost three orders of magnitude. Also, the common-emitter current gain (Ic/Ib) increased from 10-3 to approximately 0.3 for the newly designed device. However, the ON-OFF ratio of the improved germanium based GBHT has so far been rather low. Further optimizations are necessary in order to fully exploit the potential of the graphene-base heterojunction transistor.
In Memory Computing (IMC) is one of the most promising candidates
for data-intensive computing accelerators of machine learning (ML). A key ML algorithm for dimensionality reduction and classification is principal component analysis (PCA), which heavily relies on matrixvector multiplications (MVM) for which classic von Neumann architectures are not optimized. Here, we provide the experimental
demonstration of a new IMCbased PCA algorithm based on power iteration and deflation executed in a 4-kbit array of resistive switching random-access memory (RRAM). The classification accuracy of the Wisconsin Breast Cancer data set reaches 95.43%, close to floatingpoint implementation. Our simulations indicate a 250× improvement in energy efficiency compared to commercial GPUs, thus supporting IMC for energy-efficient ML in modern data-intensive computing.
We have performed different simulation experiments in relation to hardware neural networks (NN) to analyze the role of the number of synapses for different NN architectures in the network accuracy, considering different datasets. A technology that stands upon 4-kbit 1T1R ReRAM arrays, where resistive switching devices based on HfO2 dielectrics are employed, is taken as a reference. In our study, fully dense (FdNN) and convolutional neural networks (CNN) were considered, where the NN size in terms of the number of synapses and of hidden layer neurons were varied. CNNs work better when the number of synapses to be used is limited. If quantized synaptic weights are included, we observed thatNNaccuracy decreases significantly as the number of synapses is reduced; in this respect, a trade-off between the number of synapses and the NN accuracy has to be achieved. Consequently, the CNN architecture must be carefully designed; in particular, it was noticed that different datasets need specific architectures according to their complexity to achieve good results. It was shown that due to the number of variables that can be changed in the optimization of a NN hardware implementation, a specific solution has to be worked in each case in terms of synaptic weight levels, NN architecture, etc.
Resistive switching memory (RRAM) is a promising technology for highly efficient computing scenarios. RRAM arrays enabled the acceleration of neural networks for artificial intelligence and the creation of In-Memory Computing circuits. However, the arrays are affected by several issues materializing in conductance variations that might cause severe performance degradation in those applications. Among those, one is related to the drift of the low conductance states appearing immediately at the end of program and verify algorithms that are fundamental for an accurate Multi-level conductance operation. In this work, we tackle the issue by developing an Incremental Reset and Verify technique showing enhanced variability and reliability features compared with a traditional refresh-based approach.
With the rapid emergence of in-memory computing systems based on memristive technology, the integration of such memory devices in large-scale architectures is one of the main aspects to tackle.
In this work we present a study of HfO2-based memristive devices for their integration in large-scale CMOS systems, namely 200 mm wafers. The DC characteristics of single metal–insulator–metal
devices are analyzed taking under consideration device-to-device variabilities and switching properties. Furthermore, the distribution of the leakage current levels in the pristine state of the
samples are analyzed and correlated to the amount of formingless memristors found among the measured devices. Finally, the obtained results are fitted into a physic-based compact model that
enables their integration into larger-scale simulation environments.
This work reports forming free/self-rectifying resistive switching characteristics and dependency of the top electrode (TE) of a crystalline HfO2-based resistive switching memory device. In the memory cells, nonlinear bipolar resistive switching characteristics, i.e., an asymmetric current-voltage curve like the Schottky diode, was
observed. In addition, the device exhibits resistive switching behaviors without forming process, which makes it possible to switch the resistance state under ultra-low current levels of <10 nA. In addition, compared to the resistive switching of the proposed resistive switching memory devices with different TEs, the VSET was decreased when using TE with lower work function, and the height read margin was obtained in the sample with the Ni TE, covering over 56 × 56 arrays. Consequently, these results indicate that the interface control resistive switching properties in memory structures having the Schottky junction warrant the realization of selector-free resistive
switching memory cells in a high-density crossbar array.
Influence of variability on the performance of HfO2 memristor-based convolutional neural networks
(2021)
A study of convolutional neural networks (CNNs) was performed to analyze the influence of quantization and variability in the network synaptic weights. Different CNNs were considered accounting for the number of convolutional layers, size of the filters in the convolutional layer, number of neurons in the final network layers and different sets of quantization levels. The conductance levels of fabricated 1T1R structures based on HfO2 memristors were considered as reference for four or eight level quantization processes at the inference stage of the CNNs, which were previous trained with the MNIST dataset. We also included the variability of the experimental conductance levels that was found to be Gaussian distributed and was correspondingly modeled for the synaptic weight implementation.