Application of the Quasi-Static Memdiode Model in Cross-Point Arrays for Large Dataset Pattern Recognition
Open accessThe DOI matches its Crossref record; 98 of the 100 references it cites were found in a registry or an index.
Abstract
We investigate the use and performance of the quasi-static memdiode model (QMM) when incorporated into large cross-point arrays intended for pattern classification tasks. Following Chua's memristive devices theory, the QMM comprises two equations, one equation for the electron transport based on the double-diode circuit with single series resistance and a second equation for the internal memory state of the device based on the so-called logistic hysteron or memory map. Ex-situ trained memdiodes with different MNIST-like databases are used to establish the synaptic weights linking the top and bottom wire networks. The role played by the memdiode electrical parameters, wire resistance and capacitance values, image pixelation, connection schemes, signal-to-noise ratio and device-to-device variability in the classification effectiveness are investigated. The confusion matrix is used to benchmark the system performance metrics. We show that the simplicity, accuracy and robustness of the memdiode model makes it a suitable candidate for the realistic simulation of large-scale neural networks with non-idealities.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Figures

![FIGURE 2 (a) Hysteron model with logistic ridge functions 0 C (Eq. 3) and 0 (Eq. 4). is the space of feasible states S . The red thick faded line superimposed to the hysteron model indicates the trajectory of the state variable inside from an initial ( S 1 ) to a final ( S 2 ) state. The inset in the left shows the equivalent circuit model for the current equation (1) including the series resistance. The diodes are driven by the memory state of the device and one diode is activated at a time. (b) Typical I -V characteristic for a memdiode obtained via simulation of the proposed model. Superimposed blue arrows indicate the current evolution. (c) Detail of the HRS (exponential) to LRS (linear) transition, showing a few intermediate states. Linear [13], [15], [16] and sinh() -like [34] models are superimposed for comparison.](https://figures.cochat.ai/paper-figures/99ebb1d0b1804254a3a09cbce52013e5/application-of-the-quasi-static-memdiode-model-in-cross-point-arrays-for-large-figure-2.png)
![FIGURE 3 Experimental I -V loops of different materials reported in the literature fitted with the QMM model: (a) HfO 2 [43], (b) Al 2 O 3 [44], (c) MnO 3 [45], (d) CuO 2 [46], (e) La 1-x Cax MnO 3 [47] and (f) TaO X [48]. The QMM fitting parameters are shown for each case. As reference, the HRS and LRS curves are indicated in (a) and the SET and RESET points in (b). Note that in (a) a current compliance of 200 A was imposed to prevent permanent dielectric breakdown, which can be also represented by the QMM.](https://figures.cochat.ai/paper-figures/6c1c1d4b6f5e4e24985a5266543d4109/application-of-the-quasi-static-memdiode-model-in-cross-point-arrays-for-large-figure-3.png)
![FIGURE 4 (a) Schematic representation of the Write-Verify procedure for programming the CPA memdiodes. The upper waveform stands for the alternating write and verify voltage pulses, while the lower plot represents the associated conductance changes. A simplified circuit schematic for the writing connection is indicated in the upper right inset. (b) Experimental and simulation results for the reset characteristics of SiO X from UCL (Data from [55]) using the QMM model. Notice the control of the intermediate memory states. The inset shows the input signal.](https://figures.cochat.ai/paper-figures/e78b36e9052047b88c65b25737081921/application-of-the-quasi-static-memdiode-model-in-cross-point-arrays-for-large-figure-4.png)

![FIGURE 6 Samples of the three different images databases considered in this article. In all cases images are represented in 28 28 px. Pixel brightness (or intensity) is codified in 256 levels ranging from 0 (fully OFF, black) to 1 (fully ON, white). (a) MNIST database of handwritten numeric digits, (b) MNIST-F database [57] of fashion articles and (c) MNIST-K database [58] of handwritten Kanji Japanese ideograms.](https://figures.cochat.ai/paper-figures/76386b046ac548f6ae1bc7b0c800beb8/application-of-the-quasi-static-memdiode-model-in-cross-point-arrays-for-large-figure-6.png)



![FIGURE 10 (a) CPA Power consumption vs. CPA size (in number of devices). Note that Power increases as R ON reduces from model plays ( C 1C 4). As reference, reported CPA power for other works [6], [8], [14], [79], [80] is also plotted. (b) Ratio between the memdiode power and the total CPA power consumption. (c) Obtained Signal-to-Noise ratio (SNR) for different model plays ( C 1C 4) as function of the line resistance. The marker color indicates the inference accuracy for each simulation case. Note that reduced R ON in C 1 comes at the cost of a reduced SNR and thereby limits accuracy.](https://figures.cochat.ai/paper-figures/2a51e4f775d7464195d73d287d867f52/application-of-the-quasi-static-memdiode-model-in-cross-point-arrays-for-large-figure-10.png)
![FIGURE 11 (a) Impact of the wire resistance ( R W ) on the detection performance considering four different resistance windows corresponding to models plays C 1C 4. (b) Inference accuracy is plotted against the R W = R ON ratio showing a unified trend among all model plays. Confusion matrix for model C 2, with R W equal to (c) 0.1 , (d) 4.53 (16 nm [15]) and (e) 81.3 (10 nm [76]). In all cases CPAs are connected from both sides (DSC) and images are not downsized (image resolution: 28 28 px.).](https://figures.cochat.ai/paper-figures/cc97b971909546af861f1c673dd25017/application-of-the-quasi-static-memdiode-model-in-cross-point-arrays-for-large-figure-11.png)
![FIGURE 12 (a) Expected inference latency for a CPA of 1Mb (1,000 1 : 000) as function of the process feature size, considering the model proposed by Meindl et al. in [76], [87], SPICE simulations considering the QMM, and the simplified circuital model depicted by Eq. 13. (b) Circuit schematic of a RRAM cell in a CPA structure considering the associated wire parasitic resistance and capacitance. (c) Schematic representation of the simplified model of Eq. 13. (d) Estimated Latency based on Eq. 13 as function of R OFF .](https://figures.cochat.ai/paper-figures/a448bb162a5944888ed8d69b54bb4f16/application-of-the-quasi-static-memdiode-model-in-cross-point-arrays-for-large-figure-12.png)



![FIGURE 16 Performance metrics of the memdiode-based CPA simulated with the QMM, compared against the ideal Software case computed in MATLAB. 5 different databases are considered: MNIST [78], MNIST-F [57], MNIST-K [58], CIFAR-10 [94] and SVHN [95]. Each CPA is divided into 4 partitions to further reduced the impact of R W . Note that as R W tends to 0, the CPA mimics the MATLAB results.](https://figures.cochat.ai/paper-figures/7628afdcfa2c4b098fec55c0e71d1c2f/application-of-the-quasi-static-memdiode-model-in-cross-point-arrays-for-large-figure-16.png)
![FIGURE 17 The computational cost (run-time and RAM memory usage) of the memdiode-based CPA implemented with the QMM model is compared against other memristor models reported by Yakopcic et al. [34], Laiho et al. [96] and the University of Michigan [97]. (a) Simulation run-time as function of the CPA size (in number of devices) increases quadratically. (b) Simulation RAM memory usage increases linearly with the CPA size. (c) Total CPU time normalised against the case of a resistor of fixed value to represent the programmed memristor. (d) Total RAM memory usage normalised against the resistor case (linear model). In both cases simulations were performed both in HSPICE and Fast-SPICE (FineSim) environments and for different image sizes.](https://figures.cochat.ai/paper-figures/74ee6cf51e1349db8c9c6804c7fcb080/application-of-the-quasi-static-memdiode-model-in-cross-point-arrays-for-large-figure-17.png)





Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The cited source says what the claim attributes to it.
In the citing paper
“For the case of the Quasi-static Memdiode Model (QMM) considered in refs. 250,253, this is done by adjusting the control parameter λ that runs between 0 (HRS) and 1 (LRS).”
From Hardware implementation of memristor-based artificial neural networks (2024) · cites this paper as [253]
The cited source says what the claim attributes to it.
In the citing paper
“As I0 increases in Eq. 14, the I-V curve changes its shape from exponential to linear through a continuum of states, as experimentally observed for this kind of devices253.”
From Hardware implementation of memristor-based artificial neural networks (2024) · cites this paper as [253]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Gradient-based learning applied to document recognition
Lecun, Y., Bottou, L., Bengio, Y., & Haffner, P. (1998) · Proceedings of the IEEE · doi:10.1109/5.726791 · Cited by 58,805 · cited 1×
“[78].”
An Algorithm for Least-Squares Estimation of Nonlinear Parameters
Marquardt, D. W. (1963) · Journal of the Society for Industrial and Applied Mathematics · doi:10.1137/0111030 · Cited by 30,728 · cited 1×
“Moreover, although the Levenberg-Marquardt (LM) learning algorithm [64] provides the highest accuracy at the cost of the highest CPU run-time among those considered [63]–[70], the difference between the inference accuracy”
MethodLearning Multiple Layers of Features from Tiny Images
Krizhevsky, A. (2024) · doi:10.57702/zp44cu3g · Cited by 23,395
Strukov, D. B., Snider, G. S., Stewart, D. R., & Williams, R. S. (2008) · Nature · doi:10.1038/nature06932 · Cited by 11,828 · cited 2×
“both the positive and negative synaptic weights using a Verilog-A nonlinear memristor model [12].”
BackgroundMethodNumerical Methods for Unconstrained Optimization and Nonlinear Equations
Dennis, J. E., & Schnabel, R. B. (1996) · Society for Industrial and Applied Mathematics eBooks · doi:10.1137/1.9781611971200 · Cited by 7,615
Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithms
Xiao, H., Rasul, K., & Vollgraf, R. (2017) · arXiv · doi:10.48550/arxiv.1708.07747 · Cited by 6,108
Function minimization by conjugate gradients
Fletcher, R. (1964) · The Computer Journal · doi:10.1093/comjnl/7.2.149 · Cited by 4,939
Reading digits in natural images with unsupervised feature learning
Netzer, Y. (2024) · doi:10.57702/e7wwo0zk · Cited by 4,149
A scaled conjugate gradient algorithm for fast supervised learning
Møller, M. F. (1993) · Neural Networks · doi:10.1016/s0893-6080(05)80056-5 · Cited by 4,052
A direct adaptive method for faster backpropagation learning: the RPROP algorithm
Riedmiller, M., & Braun, H. (2002) · IEEE International Conference on Neural Networks · doi:10.1109/icnn.1993.298623 · Cited by 3,917
Approximate Statistical Tests for Comparing Supervised Classification Learning Algorithms
Dietterich, T. G. (1998) · Neural Computation · doi:10.1162/089976698300017197 · Cited by 3,840 · cited 1×
“In fact, for the cases with R W ≤ 1 the differences between the accuracy of the idealised network and the QMM implementation are not statistically significant [99].”
BackgroundTraining and operation of an integrated neuromorphic network based on metal-oxide memristors
Prezioso, M., Merrikh-Bayat, F., Hoskins, B. D., Adam, G. C., Likharev, K. K., & Strukov, D. B. (2015) · Nature · doi:10.1038/nature14441 · Cited by 3,001 · cited 1×
“6 and Supplementary Tables 2-7. rendering both the positive and negative elements of W M with the always positive conductance of the CPA, each synaptic weight is implemented using two memdiodes as suggested in [71], [72] resulting in two CPAs of n 2 × 10 (20 n 2 synapses).”
MethodThe conductivity of thin metallic films according to the electron theory of metals
Fuchs, K. (1938) · Mathematical Proceedings of the Cambridge Philosophical Society · doi:10.1017/s0305004100019952 · Cited by 2,276 · cited 1×
“These two effects are well-known and can be quantified using the Fuchs-Sondheimer (FS) [85] and the Mayadas-Shatzkes (MS) [86] models, revealing that for highly scaled nodes ( ∼ 5 nm) R W can be as large as ∼ 100 k (cid:127) [76].”
BackgroundMayadas, A. F., & Shatzkes, M. (1970) · Physical Review B · doi:10.1103/physrevb.1.1382 · Cited by 2,210 · cited 1×
“These two effects are well-known and can be quantified using the Fuchs-Sondheimer (FS) [85] and the Mayadas-Shatzkes (MS) [86] models, revealing that for highly scaled nodes ( ∼ 5 nm) R W can be as large as ∼ 100 k (cid:127) [76].”
BackgroundRestart procedures for the conjugate gradient method
Powell, M. J. D. (1977) · Mathematical Programming · doi:10.1007/bf01593790 · Cited by 1,717
Analogue signal and image processing with large memristor crossbars
Li, C., Hu, M., Li, Y., Jiang, H., Ge, N., Montgomery, E., Zhang, J., Song, W., Dávila, N., Graves, C. E., et al. (2017) · Nature Electronics · doi:10.1038/s41928-017-0002-z · Cited by 1,292 · cited 5×
“As reference, reported CPA power for other works [6], [8], [14], [79], [80] is also plotted.”
BackgroundMethodInfluentialFirst- and Second-Order Methods for Learning: Between Steepest Descent and Newton's Method
Battiti, R. (1992) · Neural Computation · doi:10.1162/neco.1992.4.2.141 · Cited by 1,193
Neuro-Inspired Computing With Emerging Nonvolatile Memorys
Yu, S. (2018) · Proceedings of the IEEE · doi:10.1109/jproc.2018.2790840 · Cited by 1,170 · cited 1×
“Therefore, the limitation is not in the CPA implementation but in the Neural Network itself, i.e. for this kind of datasets more complex networks are required, as shown in [90], [100], where a Deep Neural Network comprising 6 Convolutional Layers and 2 Fully Connected layers are considered.”
MethodBurr, G. W., Shelby, R. M., Sidler, S., di Nolfo, C., Jang, J., Boybat, I., Shenoy, R. S., Narayanan, P., Virwani, K., Giacometti, E. U., et al. (2015) · IEEE Transactions on Electron Devices · doi:10.1109/ted.2015.2439635 · Cited by 936 · cited 4×
“For example, both the training and testing images are resized to 8 × 8 px. in [60], 14 × 14 px. in [56] and 22 × 24 px. in [59] by using the bicubic interpolation method.”
BackgroundMethodInfluentialEfficient and self-adaptive in-situ learning in multilayer memristor neural networks
Li, C., Belkin, D., Li, Y., Yan, P., Hu, M., Ge, N., Jiang, H., Montgomery, E., Lin, P., Wang, Z., et al. (2018) · Nature Communications · doi:10.1038/s41467-018-04484-2 · Cited by 931
Resistive switching memories based on metal oxides: mechanisms, reliability and scaling
Ielmini, D. (2016) · Semiconductor Science and Technology · doi:10.1088/0268-1242/31/6/063002 · Cited by 905 · cited 1×
“This can be achieved by limiting the CF size [77], at the cost of introducing significant variability in the resistance value, especially in R OFF [56], [77].”
BackgroundTEAM: ThrEshold Adaptive Memristor Model
Kvatinsky, S., Friedman, E. G., Kolodny, A., & Weiser, U. C. (2013) · IEEE Transactions on Circuits and Systems I: Regular Papers · doi:10.1109/tcsi.2012.2215714 · Cited by 843 · cited 1×
“Last but not least, general phenomenological models such as the Yakopcic [34], TEAM [35], VTEAM [36], and Eshraghian [37] models can successfully fit certain experimental data.”
BackgroundVTEAM: A General Model for Voltage-Controlled Memristors
Kvatinsky, S., Ramadan, M., Friedman, E. G., & Kolodny, A. (2015) · IEEE Transactions on Circuits and Systems II: Express Briefs · doi:10.1109/tcsii.2015.2433536 · Cited by 838
The elusive memristor: properties of basic electrical circuits
Joglekar, Y. N., & Wolf, S. J. (2009) · European Journal of Physics · doi:10.1088/0143-0807/30/4/001 · Cited by 817 · cited 1×
“Roughly, memristor models can be classified into three groups: first, simple behavioural models [12], [29], [30] are useful in the early stages of circuit design, i.e. when a quick proof of concept is required.”
BackgroundSwitching dynamics in titanium dioxide memristive devices
Pickett, M. D., Strukov, D. B., Borghetti, J. L., Yang, J. J., Snider, G. S., Stewart, D. R., & Williams, R. S. (2009) · Journal of Applied Physics · doi:10.1063/1.3236506 · Cited by 764 · cited 1×
“Second, device specific models such as the physical-phenomenological Pickett’s [31] and the simple-phenomenological Bayat’s [32] models for TiO 2 - based MIM structures provide the highest accuracy.”
MethodHu, M., Strachan, J. P., Li, Z., Grafals, E. M., Davila, N., Graves, C., Lam, S., Ge, N., Yang, J. J., & Williams, R. S. (2016) · Proceedings of the 53rd Annual Design Automation Conference · doi:10.1145/2897937.2898010 · Cited by 686 · cited 1×
“…only positive weights, so that In the next step, the conductance matrices G M and G M (6 and 7) to be mapped into the CPAs are calculated by the linear transformation [20], [74]: where [ G min , G max ] is a selected conductance range for a linear computation in matrix-vector calculations.”
MethodComprehensive study of the resistivity of copper wires with lateral dimensions of 100 nm and smaller
Steinhögl, W., Schindler, G., Steinlesberger, G., Traving, M., & Engelhardt, M. (2005) · Journal of Applied Physics · doi:10.1063/1.1834982 · Cited by 531 · cited 2×
“3 based on the average values reported in the literature, and d is assumed to the wire width [82], [84].”
BackgroundYu, S., Gao, B., Fang, Z., Yu, H., Kang, J., & Wong, H. ‐. P. (2013) · Advanced Materials · doi:10.1002/adma.201203680 · Cited by 527 · cited 1×
“1) are nowadays in the spotlight as their properties as programmable non-volatile memory (NVM) devices have enormous potential application in fields such as artificial intelligence (AI) [1], [2] and information storage [3].”
BackgroundStorage-class memory: The next storage system technology
Freitas, R. F., & Wilcke, W. W. (2008) · IBM Journal of Research and Development · doi:10.1147/rd.524.0439 · Cited by 510
Simple formulas for two- and three-dimensional capacitances
Sakurai, T., & Tamaru, K. (1983) · IEEE Transactions on Electron Devices · doi:10.1109/t-ed.1983.21093 · Cited by 463 · cited 1×
“8 and CW is the wire capacitance per unit length calculated as in [88]:”
MethodAlteration of Cu conductivity in the size effect regime
Rossnagel, S. M., & Kuan, T. S. (2004) · Journal of Vacuum Science & Technology B: Microelectronics and Nanometer Structures Processing, Measurement, and Phenomena · doi:10.1116/1.1642639 · Cited by 457 · cited 2×
“3 based on the average values reported in the literature, and d is assumed to the wire width [82], [84].”
BackgroundA Versatile Memristor Model With Nonlinear Dopant Kinetics
Prodromakis, T., Peh, B. P., Papavassiliou, C., & Toumazou, C. (2011) · IEEE Transactions on Electron Devices · doi:10.1109/ted.2011.2158004 · Cited by 453 · cited 1×
“Roughly, memristor models can be classified into three groups: first, simple behavioural models [12], [29], [30] are useful in the early stages of circuit design, i.e. when a quick proof of concept is required.”
BackgroundDeep Learning for Classical Japanese Literature
Clanuwat, T., Bober-Irizar, M., Kitamoto, A., Lamb, A., Yamamoto, K., & Ha, D. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1812.01718 · Cited by 442 · cited 1×
“The recognition of patterns from different databases (MNIST [42], MINST-F [57] and MNIST-K [58]) is considered for benchmarking.”
BackgroundSynaptic behaviors and modeling of a metal oxide memristive device
Chang, T., Jo, S.-H., Kim, K.-H., Sheridan, P., Gaba, S., & Lu, W. (2011) · Applied Physics A · doi:10.1007/s00339-011-6296-1 · Cited by 432 · cited 1×
“…other 4 different RRAM models proposed in the literature (Yakopcic model [34], Laiho-Biolek model [96], the University of Michigan Model [97] and a linear model [15], i.e. a fixed resistance defined upon the corresponding synaptic value) were considered apart from the QMM model for…”
MethodImmunity to Device Variations in a Spiking Neural Network With Memristive Nanodevices
Querlioz, D., Bichler, O., Dollfus, P., & Gamrat, C. (2013) · IEEE Transactions on Nanotechnology · doi:10.1109/tnano.2013.2250995 · Cited by 420 · cited 1×
“…neural networks explored in the literature (Multi-layer Perceptron [56], [59], [60], Convolutional Neural Networks [61], Spike Neural Networks [62], etc., see Supplementary Table 1), the SLP allows studying and clarifying the ANN limitations caused by parasitic effects and non-idealities…”
BackgroundMemristor Crossbar-Based Neuromorphic Computing System: A Case Study
Hu, M., Li, H., Chen, Y., Wu, Q., Rose, G. S., & Linderman, R. W. (2014) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2013.2296777 · Cited by 414 · cited 5×
“reported in [1] a simulation-based case study of a CPA for character recog-”
BackgroundInfluentialSize-Dependent Resistivity in Nanoscale Interconnects
Josell, D., Brongersma, S. H., & Tőkei, Z. (2009) · Annual Review of Materials Research · doi:10.1146/annurev-matsci-082908-145415 · Cited by 286
Liang, J., & Wong, H.-S. P. (2010) · IEEE Transactions on Electron Devices · doi:10.1109/ted.2010.2062187 · Cited by 276 · cited 1×
“As reference, reported CPA power for other works [6], [8], [14], [79], [80] is also plotted.”
MethodRescuing Memristor-based Neuromorphic Design with High Defects
Liu, C., Hu, M., Strachan, J. P., & Li, H. (. (2017) · Proceedings of the 54th Annual Design Automation Conference 2017 · doi:10.1145/3061639.3062310 · Cited by 257 · cited 2×
“Software solutions allow a more systematic study and can be split into three groups: first, several authors [13], [15]–[20] opted for solving the system of coupled differential equations which arises from considering the current Kirchoff’s law at each junction in the CPA assuming that the programmed…”
BackgroundMethodGeneralized Memristive Device SPICE Model and its Application in Circuit Design
Yakopcic, C., Taha, T. M., Subramanyam, G., & Pino, R. E. (2013) · IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems · doi:10.1109/tcad.2013.2252057 · Cited by 251 · cited 4×
“For comparison purposes, a linear [13], [15], [16] (faded-thick red lines) and sinh() -like models [34] (faded-thick black lines) are also plotted in Fig.”
BackgroundMethodInfluentialChen, A. (2013) · IEEE Transactions on Electron Devices · doi:10.1109/ted.2013.2246791 · Cited by 230
Silicon Oxide (SiOx): A Promising Material for Resistance Switching?
Mehonic, A., Shluger, A. L., Gao, D., Valov, I., Miranda, E., Ielmini, D., Bricalli, A., Ambrosi, E., Li, C., Yang, J. J., et al. (2018) · Advanced Materials · doi:10.1002/adma.201801187 · Cited by 213
Self-citationBurr, G. W., Shelby, R. M., di Nolfo, C., Jang, J. W., Shenoy, R. S., Narayanan, P., Virwani, K., Giacometti, E. U., Kurdi, B., & Hwang, H. (2014) · 2014 IEEE International Electron Devices Meeting · doi:10.1109/iedm.2014.7047135 · Cited by 208
Multilevel HfO2-based RRAM devices for low-power neuromorphic networks
Milo, V., Zambelli, C., Olivo, P., Pérez, E., K. Mahadevaiah, M., G. Ossorio, O., Wenger, C., & Ielmini, D. (2019) · APL Materials · doi:10.1063/1.5108650 · Cited by 200 · cited 7×
“…mentioning that although a simpler approach than the more complex RRAM based neural networks explored in the literature (Multi-layer Perceptron [56], [59], [60], Convolutional Neural Networks [61], Spike Neural Networks [62], etc., see Supplementary Table 1), the SLP allows studying and…”
BackgroundMethodInfluentialRRAM-based synapse for neuromorphic system with pattern recognition function
Park, S., Kim, H., Choo, M., Noh, J., Sheri, A., Jung, S., Seo, K., Park, J., Kim, S., Lee, W., et al. (2012) · 2012 International Electron Devices Meeting · doi:10.1109/iedm.2012.6479016 · Cited by 191
Bistable Resistive Switching in Al[sub 2]O[sub 3] Memory Thin Films
Lin, C.-Y., Wu, C.-Y., Wu, C.-Y., Hu, C., & Tseng, T.-Y. (2007) · Journal of The Electrochemical Society · doi:10.1149/1.2750450 · Cited by 185
Reduction and IR-drop compensations techniques for reliable neuromorphic computing systems
Liu, B., Li, H. H., Chen, Y., Li, X., Huang, T., Wu, Q., & Barnell, M. D. (2014) · International Conference on Computer Aided Design · doi:10.1109/iccad.2014.7001330 · Cited by 180 · cited 1×
“While the formers are mainly a consequence of the increase of R W as the fabrication technology scales down [10], [13] which in combination with a reduced resistance window or low R ON causes a significant voltage drop across the CPA lines, the latter refers to the non-negligible current flowing…”
BackgroundVariability of resistive switching memories and its impact on crossbar array performance
Chen, A., & Lin, M.-R. (2011) · IEEE International Reliability Physics Symposium proceedings · doi:10.1109/irps.2011.5784590 · Cited by 155 · cited 1×
“The variability of the resistance states R ON and R OFF across a matrix is largely influenced by the choice of the stack’s materials (e.g., single material HfO X vs. bilayer HfO X + TaO X ) [91], [92], as well as the device scaling.”
BackgroundFault-Tolerant Training with On-Line Fault Detection for RRAM-Based Neural Computing Systems
Xia, L., Liu, M., Ning, X., Chakrabarty, K., & Wang, Y. (2017) · Proceedings of the 54th Annual Design Automation Conference 2017 · doi:10.1145/3061639.3062248 · Cited by 152
Challenges hindering memristive neuromorphic hardware from going mainstream
Adam, G. C., Khiat, A., & Prodromakis, T. (2018) · Nature Communications · doi:10.1038/s41467-018-07565-4 · Cited by 136 · cited 1×
“However, the uncontrolled high density of defects in addition to the intrinsic stochastic nature of the switching mechanism can induce a high degree of variability [89].”
BackgroundA Collective Study on Modeling and Simulation of Resistive Random Access Memory
Panda, D., Sahu, P. P., & Tseng, T. (2018) · Nanoscale Research Letters · doi:10.1186/s11671-017-2419-8 · Cited by 133 · cited 1×
“Given its importance for reliable SPICE simulations, great attention has been put on the memristor model considered, though no general consensus has been reached on which conduction mechanism and memory equation (ME) better represent the wide spectrum of memristive behaviours [26]–[28].”
BackgroundMemristive Device Fundamentals and Modeling: Applications to Circuits and Systems Simulation
Eshraghian, K., Kavehei, O., Cho, K.-R., Chappell, J. M., Iqbal, A., Al-Sarawi, S. F., & Abbott, D. (2012) · Proceedings of the IEEE · doi:10.1109/jproc.2012.2188770 · Cited by 127 · cited 1×
“Last but not least, general phenomenological models such as the Yakopcic [34], TEAM [35], VTEAM [36], and Eshraghian [37] models can successfully fit certain experimental data.”
BackgroundInterconnect opportunities for gigascale integration
MEINDL, J. D. (2003) · IEEE Micro · doi:10.1109/mm.2003.1209464 · Cited by 123 · cited 3×
“According to Liang et al . and Meindl [76], [87] the wire latency ( τ ) can be modelled as in Eq.”
BackgroundTruong, S. N., & Min, K.-S. (2014) · JSTS:Journal of Semiconductor Technology and Science · doi:10.5573/jsts.2014.14.3.356 · Cited by 118 · cited 5×
“As reference, reported CPA power for other works [6], [8], [14], [79], [80] is also plotted.”
BackgroundMethodInfluentialMemTorch: An Open-source Simulation Framework for Memristive Deep Learning Systems
Lammie, C., Xiang, W., Linares-Barranco, B., & Azghadi, M. R. (2022) · Neurocomputing · doi:10.1016/j.neucom.2022.02.043 · Cited by 107 · cited 1×
“Second, Python-based approaches and similar [21], [22] allow incorporating realistic RRAM models, but normally ignore the parasitic CPA effects or do not account for the peripheral circuitry.”
MethodNeuroinspired unsupervised learning and pruning with subquantum CBRAM arrays
Shi, Y., Nguyen, L., Oh, S., Liu, X., Koushan, F., Jameson, J. R., & Kuzum, D. (2018) · Nature Communications · doi:10.1038/s41467-018-07682-0 · Cited by 106 · cited 2×
“These variables are: the memristor resistance window ( R ON and R OFF ) [14], [56], [59], [60], [75], the wire resistance of the CPA intercon-nections ( R W ) [15], [76] and the representation size of the dataset images ( n × n pixels) [56], [59], [60].”
BackgroundDevice and SPICE modeling of RRAM devices
Sheridan, P., Kim, K.-H., Gaba, S., Chang, T., Chen, L., & Lu, W. (2011) · Nanoscale · doi:10.1039/c1nr10557d · Cited by 98 · cited 1×
“Given its importance for reliable SPICE simulations, great attention has been put on the memristor model considered, though no general consensus has been reached on which conduction mechanism and memory equation (ME) better represent the wide spectrum of memristive behaviours [26]–[28].”
BackgroundSynaptic electronics and neuromorphic computing
Upadhyay, N. K., Joshi, S., & Yang, J. J. (2016) · Science China Information Sciences · doi:10.1007/s11432-016-5565-1 · Cited by 98 · cited 2×
“These features allow CPAs to solve a number of computationally intensive specific AI tasks such as the classification of a variety of patterns (sounds, images, electrocardiograms, etc.) with a lower energy consumption than conventional Von Neumann systems [4].”
BackgroundMethod8-Layers 3D vertical RRAM with excellent scalability towards storage class memory applications
Luo, Q., Xu, X., Gong, T., Lv, H., Dong, D., Ma, H., Yuan, P., Gao, J., Liu, J., Yu, Z., et al. (2017) · Technical digest - International Electron Devices Meeting/Technical digest · doi:10.1109/iedm.2017.8268315 · Cited by 97 · cited 1×
“The variability of the resistance states R ON and R OFF across a matrix is largely influenced by the choice of the stack’s materials (e.g., single material HfO X vs. bilayer HfO X + TaO X ) [91], [92], as well as the device scaling.”
BackgroundCross point arrays of 8 nm × 8 nm memristive devices fabricated with nanoimprint lithography
Pi, S., Lin, P., & Xia, Q. (2013) · Journal of Vacuum Science & Technology B Nanotechnology and Microelectronics Materials Processing Measurement and Phenomena · doi:10.1116/1.4827021 · Cited by 89
Truong, S. N., Ham, S.-J., & Min, K. (2014) · Nanoscale Research Letters · doi:10.1186/1556-276x-9-629 · Cited by 82
Cross-point phase change memory with 4F 2 cell size driven by low-contact-resistivity poly-Si diode
Sasago, Kinoshita, Morikawa, Kurotsuchi, Hanzawa, Mine, Shima, Fujisaki, Kume, Moriya, et al. (2009) · Symposium on VLSI Technology · Cited by 76
Memristor crossbar based hardware realization of BSB recall function
Hu, M., Li, H. H., Wu, Q., Rose, G. S., & Chen, Y. (2012) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn.2012.6252563 · Cited by 73 · cited 1×
“6 and Supplementary Tables 2-7. rendering both the positive and negative elements of W M with the always positive conductance of the CPA, each synaptic weight is implemented using two memdiodes as suggested in [71], [72] resulting in two CPAs of n 2 × 10 (20 n 2 synapses).”
MethodLiang, J., Yeh, S., Wong, S. S., & Wong, H.-S. P. (2013) · ACM Journal on Emerging Technologies in Computing Systems · doi:10.1145/2422094.2422103 · Cited by 71 · cited 10×
“An aspect ratio of 1 is assumed ( W equal to the feature size) with a barrier thickness of 2nm on each side of the wire [76].”
BackgroundResultInfluentialSimulation of Inference Accuracy Using Realistic RRAM Devices
Mehonić, A., Joksas, D., Ng, W. H., Buckwell, M., & Kenyon, A. J. (2019) · Frontiers in Neuroscience · doi:10.3389/fnins.2019.00593 · Cited by 66 · cited 1×
“Second, Python-based approaches and similar [21], [22] allow incorporating realistic RRAM models, but normally ignore the parasitic CPA effects or do not account for the peripheral circuitry.”
MethodEffect of Electrode Materials on AlN-Based Bipolar and Complementary Resistive Switching
Chen, C., Gao, S., Tang, G., Fu, H., Wang, G., Song, C., Zeng, F., & Pan, F. (2013) · ACS Applied Materials & Interfaces · doi:10.1021/am303128h · Cited by 65 · cited 8×
“Given that RRAM devices normally present a linear-exponential characteristic [81] in the HRS regime, this may lead to errors in the computation of the cell’s current when using a simple linear model for the RRAM device [13], [15], [16] as illustrated in Fig.”
BackgroundMethodInfluentialPhenomenological modeling of memristive devices
Bayat, F. M., Hoskins, B. D., & Strukov, D. B. (2015) · Applied Physics A · doi:10.1007/s00339-015-8993-7 · Cited by 62 · cited 1×
“Second, device specific models such as the physical-phenomenological Pickett’s [31] and the simple-phenomenological Bayat’s [32] models for TiO 2 - based MIM structures provide the highest accuracy.”
MethodCompact Model for the Major and Minor Hysteretic I–V Loops in Nonlinear Memristive Devices
Miranda, E. A. (2015) · IEEE Transactions on Nanotechnology · doi:10.1109/tnano.2015.2455235 · Cited by 59 · cited 4×
“From the modelling viewpoint, the compact model originally proposed by Miranda in [39] and later extended by Patterson et al. in [40] is able to describe these major (LRS) and minor (HRS) I -”
MethodInfluentialSelf-citationHandling Stuck-at-Fault Defects Using Matrix Transformation for Robust Inference of DNNs
Zhang, B., Uysal, N., Fan, D., & Ewetz, R. (2019) · IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems · doi:10.1109/tcad.2019.2944582 · Cited by 53
Yakopcic, C., Taha, T. M., Subramanyam, G., & Pino, R. E. (2012) · Springer eBooks · doi:10.1007/978-94-007-4491-2_12 · Cited by 51 · cited 1×
“reduced times, but it seems unable to handle step functions as well as if-statements in the memristor model [25].”
BackgroundHandling stuck-at-faults in memristor crossbar arrays using matrix transformations
Zhang, B., Uysal, N., Fan, D., & Ewetz, R. (2019) · Asia and South Pacific Design Automation Conference (ASP-DAC) · doi:10.1145/3287624.3287707 · Cited by 48
Convolutional Neural Networks Based on RRAM Devices for Image Recognition and Online Learning Tasks
Dong, Z., Zhou, Z., Li, Z., Liu, C., Huang, P., Liu, L., Liu, X., & Kang, J. (2019) · IEEE Transactions on Electron Devices · doi:10.1109/ted.2018.2882779 · Cited by 47 · cited 2×
“As the variability is normally higher in HRS than in LRS [61], they swept independently, resulting in the accuracy maps illustrated in Figs.”
BackgroundMethodA Memristor Crossbar Based Computing Engine Optimized for High Speed and Accuracy
Liu, C., Yang, Q. X., Yan, B., Yang, J., Du, X., Zhu, W., Jiang, H., Wu, Q., Barnell, M. D., & Li, H. H. (2016) · IEEE Computer Society Annual Symposium on VLSI · doi:10.1109/isvlsi.2016.46 · Cited by 46 · cited 1×
“As reference, reported CPA power for other works [6], [8], [14], [79], [80] is also plotted.”
MethodA general memristor model and its applications in programmable analog circuits
Zha, J., Huang, H., Huang, T., Cao, J., Alsaedi, A., & Alsaadi, F. E. (2017) · Neurocomputing · doi:10.1016/j.neucom.2017.04.057 · Cited by 44
Truong, S. N., Shin, S., Byeon, S.-D., Song, J.-S., & Min, K. (2015) · IEEE Transactions on Nanotechnology · doi:10.1109/tnano.2015.2473666 · Cited by 42 · cited 2×
“As reference, reported CPA power for other works [6], [8], [14], [79], [80] is also plotted.”
MethodReliable Modeling of Ideal Generic Memristors via State-Space Transformation
Biolek, Z., Biolek, D., Biolková, V., & Kolka, Z. (2015) · Radioengineering · doi:10.13164/re.2015.0393 · Cited by 42 · cited 1×
“Nevertheless, the latter two groups rely on various internal equations or the introduction of an artificial window function in the ME (commonly used for modelling the SET/RESET transitions) which pose serious mathematical drawbacks causing convergence problems [33], [38].”
BackgroundMask Technique for Fast and Efficient Training of Binary Resistive Crossbar Arrays
Fouda, M. E., Lee, S., Lee, J., Eltawil, A., & Kurdahi, F. (2019) · IEEE Transactions on Nanotechnology · doi:10.1109/tnano.2019.2927493 · Cited by 41 · cited 3×
“…vertical interconnects used to connect the outputs of the vertical CPA partitions may be placed under the partitioned struc-FIGURE ture, as well as the analogue sensing electronics, allowing the partitioned CPA to maintain a similar area consumption than the original non-partitioned case [73].”
MethodVoltage-Driven Hysteresis Model for Resistive Switching: SPICE Modeling and Circuit Applications
Patterson, G. A., Suñé, J., & Miranda, E. A. (2017) · IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems · doi:10.1109/tcad.2017.2756561 · Cited by 37 · cited 4×
“From the modelling viewpoint, the compact model originally proposed by Miranda in [39] and later extended by Patterson et al. in [40] is able to describe these major (LRS) and minor (HRS) I -”
MethodInfluentialSelf-citationPetzold, S., Miranda, E. A., Sharath, S. U., Muñoz-Gorriz, J., Vogel, T., Piros, E., Kaiser, N., Eilhardt, R., Zintler, A., Molina‐Luna, L., et al. (2019) · Journal of Applied Physics · doi:10.1063/1.5094864 · Cited by 32 · cited 1×
“Remarkably, the QMM has been explored so far as a single device or as simple series/anti-series/parallel/anti-parallel connections (just two devices) [39]–[41], and its application to the case of large CPAs for pattern recognition tasks is still to be addressed.”
Self-citationTheory of Computational Complexity
Du, D., & Ko, K. (2014) · Wiley eBooks · doi:10.1002/9781118595091 · Cited by 29
Modeling of TiO2memristor: from analytic to numerical analyses
Biolek, D., Biolek, Z., Biolková, V., & Kolka, Z. (2014) · Semiconductor Science and Technology · doi:10.1088/0268-1242/29/12/125008 · Cited by 27 · cited 2×
“However, given their high computational cost, they may not be suitable in a scenario involving a large number of devices [33].”
BackgroundComplementary resistive switching in single sandwich structure for crossbar memory arrays
Duan, W., Wang, J. B., Zhong, X. L., Song, H., & Li, B. (2016) · Journal of Applied Physics · doi:10.1063/1.4961222 · Cited by 25
Ham, S.-J., Mo, H.-S., & Min, K. (2013) · IEEE Transactions on Nanotechnology · doi:10.1109/tnano.2013.2274529 · Cited by 24 · cited 2×
“Such applications have been extensively studied in previous works [1], [6]–[11] considering various CPA architectures as well as different memristor models.”
BackgroundHan, R., Huang, P. C., Zhao, Y., Cui, X., Liu, X., & Kang, J. (2018) · Science China Information Sciences · doi:10.1007/s11432-018-9555-8 · Cited by 24 · cited 5×
“Given that RRAM devices normally present a linear-exponential characteristic [81] in the HRS regime, this may lead to errors in the computation of the cell’s current when using a simple linear model for the RRAM device [13], [15], [16] as illustrated in Fig.”
BackgroundMethodInfluentialLee, Y. K., Jeon, J. W., Park, E.-S., Yoo, C., Kim, W., Ha, M., & Hwang, C. S. (2019) · Micromachines · doi:10.3390/mi10050306 · Cited by 22 · cited 9×
“…4 different RRAM models proposed in the literature (Yakopcic model [34], Laiho-Biolek model [96], the University of Michigan Model [97] and a linear model [15], i.e. a fixed resistance defined upon the corresponding synaptic value) were considered apart from the QMM model for comparison purposes.”
BackgroundMethodInfluentialEfficacy of memristive crossbars for neuromorphic processors
Yakopcic, C., Hasan, R., Taha, T. M., McLean, M. R., & Palmer, D. (2014) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn.2014.6889807 · Cited by 19 · cited 1×
“However, this approach is constrained to the limitations of the memristor model considered and to the size of the memristor-based CPA given the high computational requirements [23].”
MethodLee, A. R., Bae, Y. C., Im, H. S., & Hong, J. P. (2013) · Applied Surface Science · doi:10.1016/j.apsusc.2013.02.100 · Cited by 18 · cited 1×
“The movement is originated by the application of an external electrical stimulus, current or voltage [49]–[52].”
BackgroundAguirre, F. L., Rodriguez-Fernandez, A., Pazos, S. M., Suñé, J., Miranda, E. A., & Palumbo, F. (2019) · IEEE Transactions on Electron Devices · doi:10.1109/ted.2019.2922555 · Cited by 15 · cited 2×
“V relationship, while the completion of the CF leads to the low resistance state (LRS), which often exhibits a linear current or voltage dependence [48], [53].”
BackgroundSelf-citationResistive switching characteristics of TiN/MnO2/Pt memory devices
Yang, M., Park, J., Ko, T. K., & Lee, J.-K. (2010) · physica status solidi (RRL) - Rapid Research Letters · doi:10.1002/pssr.201004213 · Cited by 13 · cited 2×
“V loops of different materials reported in the literature fitted with the QMM model: (a) HfO 2 [43], (b) Al 2 O 3 [44], (c) MnO 3 [45], (d) CuO 2 [46], (e) La 1-x Ca x MnO 3 [47] and (f) TaO X [48].”
BackgroundMethodEffect of Device Variation on Mapping Binary Neural Network to Memristor Crossbar Array
Yi, W., Kim, Y., & Kim, J. (2019) · Design, Automation, and Test in Europe · doi:10.23919/date.2019.8714817 · Cited by 13 · cited 3×
“Along with the read margin reduction caused by the limited resistance window ( R ON / R OFF ) and interconnection line resistance ( R W ), such variability in the memristive structures also increases the possibility of a deviation of the weighted sum from the target value [90].”
BackgroundMethodMiranda, E. A., Acevedo, W. R., Rubi, D., Lüders, U., Granell, P., Suñé, J., & Lévy, P. (2017) · Journal of Applied Physics · doi:10.1063/1.4984051 · Cited by 12 · cited 2×
“Experimental I-V loops of different materials reported in the literature fitted with the QMM model: (a) HfO2 [43], (b) Al2O3 [44], (c) MnO3 [45], (d) CuO2 [46], (e) La1-xCaxMnO3 [47] and (f) TaOX [48].”
BackgroundMethodSelf-citationChoi, W., Moon, K., Kwak, M., Sung, C., Lee, J., Song, J., Park, J., Chekol, S. A., & Hwang, H. (2019) · Solid-State Electronics · doi:10.1016/j.sse.2018.12.018 · Cited by 12 · cited 1×
“Second, the sinh() -like model, requires the simultaneous modification of multiple parameters to mimic the smooth linear-exponential to linear transition or even separate expressions for the HRS and LRS regimes [16], [54].”
BackgroundFröhlich, K., Kundrata, I., Blaho, M., Precner, M., Ťapajna, M., Klimo, M., Šuch, O., & Škvarek, O. (2018) · Journal of Applied Physics · doi:10.1063/1.5025802 · Cited by 11 · cited 4×
“V relationship, while the completion of the CF leads to the low resistance state (LRS), which often exhibits a linear current or voltage dependence [48], [53].”
BackgroundMethodInfluentialMemristive synapses are becoming reality
Laiho, M., Lehtonen, E., Russel, A., et al. (2010) · Research Explorer (The University of Manchester) · doi:10.2417/1201011.003396 · Cited by 9
Nonvolatile State Identification and NVM SPICE
Yu, H., & Wang, Y. (2014) · Springer eBooks · doi:10.1007/978-1-4939-0551-5_3 · Cited by 6
Blasco, J., Jančovič, P., Fröhlich, K., Suñé, J., & Miranda, E. A. (2014) · Journal of Vacuum Science & Technology B Nanotechnology and Microelectronics Materials Processing Measurement and Phenomena · doi:10.1116/1.4900599 · Cited by 6
Self-citationYang, M., Wang, H., Ma, X., Gao, H., & Yue, H. (2017) · Journal of Vacuum Science & Technology B Nanotechnology and Microelectronics Materials Processing Measurement and Phenomena · doi:10.1116/1.4983193 · Cited by 5
On the properties of conducting filament in ReRAM
Lian, X., Lanza, M., Rodriguez-Fernandez, A., Miranda, E. A., & Suñé, J. (2014) · IEEE International Conference on Solid-State and Integrated Circuit Technology · doi:10.1109/icsict.2014.7021484 · Cited by 4 · cited 1×
“Given that RRAM devices normally present a linear-exponential characteristic [81] in the HRS regime, this may lead to errors in the computation of the cell’s current when using a simple linear model for the RRAM device [13], [15], [16] as illustrated in Fig.”
MethodSelf-citation
Research works printed but not linked: 3
- 2010, Resistance Random Access Memory (RRAM)
- lecun, 1998, MNIST Handwritten Digit Database
- hagan, 2014, Neural Network Design
Cited by
Hardware implementation of memristor-based artificial neural networks
Aguirre, F., Sebastian, A., Le Gallo, M., Song, W., Wang, T., Yang, J. J., Lu, W., Chang, M.-F., Ielmini, D., Yang, Y., et al. (2024) · Nature Communications · doi:10.1038/s41467-024-45670-9 · Cited by 564
“250,253, this is done by adjusting the control parameter λ that runs between 0 (HRS) and 1 (LRS).”
InfluentialSelf-citationIntegrated Memory Devices Based on 2D Materials
Xue, F., Zhang, C., Ma, Y., Wen, Y., He, X., Yu, B., & Zhang, X. (2022) · Advanced Materials · doi:10.1002/adma.202201880 · Cited by 98
Solution-processed memristors: performance and reliability
Pazos, S. M., Xu, X., Guo, T., Zhu, K., Alshareef, H. N., & Lanza, M. (2024) · Nature Reviews Materials · doi:10.1038/s41578-024-00661-6 · Cited by 110
“For deeply scaled crossbar arrays, one needs to bear in mind that crossbar architectures bring many associated nonideali-ties, including wire resistance, sneak-paths and failed devices (stuck-at-faults) 100 .”
Self-citationYang, J., Yoon, A., Lee, D., Song, S., Jung, I. J., Lim, D., Jeong, H., Lee, Z., Lanza, M., & Kwon, S. (2023) · Advanced Functional Materials · doi:10.1002/adfm.202309455 · Cited by 62
SPICE Implementation of the Dynamic Memdiode Model for Bipolar Resistive Switching Devices
Aguirre, F. L., Suñé, J., & Miranda, E. A. (2022) · Micromachines · doi:10.3390/mi13020330 · Cited by 59
Self-citation
Metadata
- Authors
- Aguirre, F. L., Pazos, S. M., Palumbo, F., Sune, J., & Miranda, E.
- Affiliations
- Universitat Autònoma de Barcelona (ES)
- National Technological University (AR)
- Venue
- IEEE Access
- Published
- 2020
- Volume
- 8
- Pages
- 202174-202193
- ISSN
- 2169-3536 (ISSN-L)
- OpenAlex
- W3096521724
- Cited by
- 36
- References
- 100
- Citations by year
- 2026: 6 · 2025: 6 · 2024: 6 · 2023: 6 · 2022: 7 · 2021: 4 · 2020: 1
- FWCI
- 1.54
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Locations
- IEEE Access (publishedVersion, cc-by)
- Dipòsit Digital de Documents de la UAB (Universitat Autònoma de Barcelona) (submittedVersion, cc-by)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- CONICET Digital (CONICET) (submittedVersion, cc-by)
- Greater South Information System
- Funders
- Universidad Tecnolgica Nacional (PID-UTN CCUTIBA4764TC,, PID-UTN CCUTNBA5182, PID-UTN CCUTNBA6615, PID-UTN EIUTIBA4395TC3, PID-UTN MATUNBA4936)
- Electronic Components and Systems for European Leadership (WAKeMeUP 783176)
- Consejo Nacional de Investigaciones Cientficas y Tcnicas (PIP-11220130100077CO)
- Ministerio de Ciencia Tecnologa e Innovacin Productiva (PICT 2013/1210, PICT 2016/0579, PME 2015-0196)
- Ministerio de Ciencia e Innovacin (TEC2017-84321-C4-4-R)
- Topics
- Physical Sciences › Engineering › Electrical and Electronic Engineering › Advanced Memory and Neural Computing
- Physical Sciences › Computer Science › Artificial Intelligence › Neural Networks and Applications
- Physical Sciences › Engineering › Electrical and Electronic Engineering › Ferroelectric and Negative Capacitance Devices
- Keywords
- cross-point array, in-memory computing, pattern classification, MNIST, synaptic weights, device-to-device variation, neuromorphic hardware
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.