Stuck-at Fault Tolerance in RRAM Computing Systems
The DOI matches its Crossref record; 30 of the 31 references it cites were found in a registry or an index.
Abstract
Emerging metal-oxide resistive switching random-access memory (RRAM) devices and RRAM crossbars have demonstrated their potential in boosting the speed and energy-efficiency of analog matrix-vector multiplication. However, due to the immature fabrication technology, commonly occurring Stuck-At-Faults (SAFs) seriously degrade the computational accuracy of an RRAM-based computing system (RCS). In this paper, we present a fault-tolerant framework for RCS. A mapping algorithm with inner fault tolerance is proposed to convert matrix parameters into RRAM conductances in RCS and tolerate SAFs by fully exploring the available mapping space. Two baseline redundancy schemes are proposed to ensure that RCS is effective when the percentage of faulty RRAM cells is high. To reduce the number of redundant RRAM cells when the SAFs follow a non-uniform distribution or an unknown distribution, a distribution-aware redundancy scheme and a re-configurable redundancy scheme are proposed to provide dynamic fault tolerance. Simulation results show that, the baseline redundancy schemes can improve the recognition accuracy of the MNIST data set to almost the same as the RRAM-fault-free case, with an energy overhead of approximately 30%. When SAFs follow a non-uniform and an unknown distribution, the distribution-aware and re-configurable schemes can reduce the number of redundant RRAM cells from more than 200% to less than 40% and 60%, respectively, without reducing the recognition accuracy.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
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?
In the citing paper
“However, grounding the bitlines through a resistor might alter the potential applied to the bitline, which will no longer be 0 volts, adding variability and thus altering the read over the sensing resistor100,112.”
From Hardware implementation of memristor-based artificial neural networks (2024) · cites this paper as [112]
Citations
Cites
Wong, H.-S. P., Lee, H.-Y., Yu, S., Chen, Y.-S., Wu, Y., Chen, P.-S., Lee, B., Chen, F. T., & Tsai, M.-J. (2012) · Proceedings of the IEEE · doi:10.1109/jproc.2012.2190369 · Cited by 2,757 · cited 1×
“RRAM is an emerging passive two-ports device based on metal oxide materials such as HfOx , TiOx and WOx [17], which has been widely used as non-volatile memory.”
BackgroundISAAC: a convolutional neural network accelerator with in-situ analog arithmetic in crossbars
Shafiee, A., Nag, A., Muralimanohar, N., Balasubramonian, R., Strachan, J. P., Hu, M., Williams, R. S., & Srikumar, V. (2016) · ACM SIGARCH Computer Architecture News · doi:10.1145/3007787.3001139 · Cited by 1,536 · cited 2×
“, RRAM-based dot-product operator [4], [20], RRAM-based fully-connected Neural Networks (NN) [21], [22], and RRAM-based convolutional neural networks [15], [23].”
MethodChi, P., Li, S., Xu, C., Zhang, T., Zhao, J., Liu, Y., Wang, Y., & Xie, Y. (2016) · ACM SIGARCH Computer Architecture News · doi:10.1145/3007787.3001140 · Cited by 1,231 · cited 2×
“The most attractive feature of RRAM devices is that the crossbar structure of RRAM array can be used to perform matrixvector multiplication, which efficiently merges the memory and computation together [18] and reduces the computation complexity from O(n2) to O(1) [19].”
Self-citationHu, 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×
“, RRAM-based dot-product operator [4], [20], RRAM-based fully-connected Neural Networks (NN) [21], [22], and RRAM-based convolutional neural networks [15], [23].”
Memristor 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 3×
“Previously proposed mapping algorithms, which convert matrix parameters into RRAM conductances of the RCS, use simplified assumptions and are targeted at ideal RRAM devices without SAFs [4], [13], [14].”
BackgroundMethodReal-time EEG analysis with subject-specific spatial patterns for a brain-computer interface (BCI)
Guger, C., Ramoser, H., & Pfurtscheller, G. (2000) · IEEE Transactions on Rehabilitation Engineering · doi:10.1109/86.895947 · Cited by 383 · cited 1×
“2776980 GPU, and FPGA, cannot meet the requirements of processing data in real-time with high energy efficiency [1].”
BackgroundRRAM Defect Modeling and Failure Analysis Based on March Test and a Novel Squeeze-Search Scheme
Chen, C.-Y., Shih, H.-C., Wu, C.-W., Lin, C.-H., Chiu, P.-F., Sheu, S.-S., & Chen, F. T. (2015) · IEEE Transactions on Computers · doi:10.1109/tc.2014.12 · Cited by 261 · cited 9×
“For example, researchers reported that only 63% of the RRAM cells are fault-free for a fabricated 4-Mb HfO2-based RRAM test chip, and SAFs appear rather frequently - about 10% of the RRAM cells in the chip contain SAFs [5].”
BackgroundMethodInfluentialRescuing 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 3×
“In future work, we will combine the proposed hardwarelevel schemes with algorithm-level methods [8], [9], [27] to further improve the fault tolerance of RCS.”
MethodAccelerator-friendly neural-network training: Learning variations and defects in RRAM crossbar
Chen, L., Li, J., Chen, Y., Deng, Q., Shen, J., Liang, X., & Jiang, L. (2017) · Design, Automation & Test in Europe Conference & Exhibition (DATE), 2017 · doi:10.23919/date.2017.7926952 · Cited by 245 · cited 4×
“In future work, we will combine the proposed hardwarelevel schemes with algorithm-level methods [8], [9], [27] to further improve the fault tolerance of RCS.”
BackgroundMethodInfluentialTechnological Exploration of RRAM Crossbar Array for Matrix-Vector Multiplication
Xia, L., Gu, P., Li, B., Tang, T., Yin, X., Huangfu, W., Yu, S., Cao, Y., Wang, Y., & Yang, H. (2016) · Journal of Computer Science and Technology · doi:10.1007/s11390-016-1608-8 · Cited by 183 · cited 6×
“In prior work, after an RCS is fabricated, the parameters in the computed matrices are first mapped to the conductances of RRAM cells [4], [14].”
BackgroundMethodSelf-citation
Research works printed but not linked: 2
- lecun, 0, The MNIST Database of Handwritten Digits
- prakash, 2013, Improved resistance memory characteristics and switching mechanism using TiN electrode on TaOx/W structure, Proc Winter Simulation Conf, 136
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
“There are three widely used sensing modes for the output voltages 112 .”
Reliability of analog resistive switching memory for neuromorphic computing
Zhao, M., Gao, B., Tang, J., Qian, H., & Wu, H. (2020) · Applied Physics Reviews · doi:10.1063/1.5124915 · Cited by 349
Mehonic, A., Sebastian, A., Rajendran, B., Simeone, O., Vasilaki, E., & Kenyon, A. J. (2020) · Advanced Intelligent Systems · doi:10.1002/aisy.202000085 · Cited by 331
Committee machines—a universal method to deal with non-idealities in memristor-based neural networks
Joksas, D., Freitas, P., Chai, Z., Ng, W. H., Buckwell, M., Li, C., Zhang, W., Xia, Q., Kenyon, A. J., & Mehonić, A. (2020) · RePEc: Research Papers in Economics · Cited by 237
He, Z., Lin, J., Ewetz, R., Yuan, J.-S., & Fan, D. (2019) · Proceedings of the 56th Annual Design Automation Conference 2019 · doi:10.1145/3316781.3317870 · Cited by 149
“Many previous studies [3, 17] only discuss the effect of SAF on NN inference, using a simple 2-layers fully-connected network on MNIST, where the results show that SAF indeed affects the accuracy but with a relatively low degradation.”
BackgroundMethodCommittee machines—a universal method to deal with non-idealities in memristor-based neural networks
Joksas, D., Freitas, P., Chai, Z., Ng, W. H., Buckwell, M., Li, C., Zhang, W. D., Xia, Q., Kenyon, A. J., & Mehonić, A. (2020) · Nature Communications · doi:10.1038/s41467-020-18098-0 · Cited by 267
“…ALD-TiN buffer layer is introduced B. Govoreanu (2013) 33 Device-to-device variability Al 2 O 3 /TiO 2 (VMCO) RRAM Non-fi lamentary RRAM is adopted A. J. Kenyon (2019) 34 Device-to-device variability SiO x RRAM The roughness of bottom electrodes is increased L. Xia (2017) 14 Faulty devices –”
MethodTestability and Dependability of AI Hardware: Survey, Trends, Challenges, and Perspectives
Su, F., Liu, C., & Stratigopoulos, H. (2023) · IEEE Design and Test · doi:10.1109/mdat.2023.3241116 · Cited by 97
“In [168], spatial redundancy-based fault-tolerant schemes are proposed for yield loss recovery of memristor crossbars.”
BackgroundKim, K., Song, M. S., Hwang, H., Hwang, S., & Kim, H. (2024) · Frontiers in Neuroscience · doi:10.3389/fnins.2024.1279708 · Cited by 81
TIME: A Training-in-Memory Architecture for RRAM-Based Deep Neural Networks
Cheng, M., Xia, L., Zhu, Z., Cai, Y., Xie, Y., Wang, Y., & Yang, H. (2018) · IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems · doi:10.1109/tcad.2018.2824304 · Cited by 75
“For this problem, our system can adopt the fault-tolerant method in [38] and [39] to reduce the negative effects.”
MethodSelf-citationStuck-at-Fault Tolerant Schemes for Memristor Crossbar Array-Based Neural Networks
Yeo, I., Chu, M., Gi, S.-G., Hwang, H., & Lee, B.-G. (2019) · IEEE Transactions on Electron Devices · doi:10.1109/ted.2019.2914460 · Cited by 55
“In previous works, training algorithms [16]–[19], which require redundant hardware either to replace faulty memristors with the functional ones [16], [19] or to correct errors using”
Multi-Objective Optimization of ReRAM Crossbars for Robust DNN Inferencing under Stochastic Noise
Yang, X., Belakaria, S., Joardar, B. K., Yang, H., Doppa, J. R., Pande, P. P., Chakrabarty, K., & Li, H. H. (2021) · International Conference on Computer Aided Design · doi:10.1109/iccad51958.2021.9643444 · Cited by 59
Self-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
“Nevertheless, the use of additional rows results in additional overhead [25].”
BackgroundMethodInfluentialHuang, C., Xu, N., Qiu, K., Zhu, Y., Ma, D., & Fang, L. (2021) · IEEE Journal of the Electron Devices Society · doi:10.1109/jeds.2021.3093478 · Cited by 50
Seizure Detection and Prediction by Parallel Memristive Convolutional Neural Networks
Li, C., Lammie, C., Dong, X., Amirsoleimani, A., Azghadi, M. R., & Genov, R. (2022) · IEEE Transactions on Biomedical Circuits and Systems · doi:10.1109/tbcas.2022.3185584 · Cited by 55
“proposed a mapping algorithm with inner fault tolerance to leverage the differential mapping scheme of crossbar arrays to tolerate faults [77].”
BackgroundAn FPGA-Based Hardware Emulator for Neuromorphic Chip With RRAM
Luo, T., Wang, X., Qu, C., Lee, M. K. F., Tang, W. T., Wong, W., & Goh, R. S. M. (2018) · IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems · doi:10.1109/tcad.2018.2889670 · Cited by 48
“Following previous works [15], [23], for stuck-at-short defects, we varied the occurrence probability from 0% to 1%.”
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
“Note that the first row in the figure is equivalent to the technique of programming parallel rows used in [21].”
BackgroundMethodResultInfluentialIR-QNN Framework: An IR Drop-Aware Offline Training of Quantized Crossbar Arrays
Fouda, M. E., Lee, S., Lee, J., Kim, G. H., Kurdahi, F. J., & Eltawil, A. M. (2020) · IEEE Access · doi:10.1109/access.2020.3044652 · Cited by 44
“the network with the full knowledge of the SAF devices’ locations, especially that the reported SAF percentages are less than 20% [44], [45].”
BackgroundSun, H., Zhu, Z., Cai, Y., Chen, X., Wang, Y., & Yang, H. (2020) · Asia and South Pacific Design Automation Conference (ASP-DAC) · doi:10.1109/asp-dac47756.2020.9045192 · Cited by 41
“According to the Trans-impedance amplifier (TIA) sensing model [14], the energy consumption of each MVM result in PIM can be calculated by Equation 7. where E j denotes the energy consumption for generating I j , and t represents the clock cycle of the analog computing.”
MethodSelf-citationNonideality‐Aware Training for Accurate and Robust Low‐Power Memristive Neural Networks
Joksas, D., Wang, E., Barmpatsalos, N., Ng, W. H., Kenyon, A. J., Constantinides, G. A., & Mehonic, A. (2022) · Advanced Science · doi:10.1002/advs.202105784 · Cited by 39
“For general applications, mapping or redundancy schemes can be used to mitigate the effects of faulty-devices [27] or line resistance [28].”
BackgroundFault tolerance in neuromorphic computing systems
Liu, M., Xia, L., Wang, Y., & Chakrabarty, K. (2019) · Asia and South Pacific Design Automation Conference (ASP-DAC) · doi:10.1145/3287624.3288743 · Cited by 35
“A reconfigurable redundancy scheme was then proposed to tolerate dynamic faults [33].”
MethodSelf-citationOxide-based filamentary RRAM for deep learning
Zhang, Y., Huang, P. C., Gao, B., Kang, J., & Wu, H. (2020) · Journal of Physics D Applied Physics · doi:10.1088/1361-6463/abc5e7 · Cited by 39
“The accuracy degradation caused by a stuck device can be solved by optimizing the weight mapping method, such as introducing redundancy RRAM rows [83] or remapping the synaptic weight considering the distribution of stuck devices [84].”
MethodResistive Neural Hardware Accelerators
Smagulova, K., Fouda, M. E., Kurdahi, F. J., Salama, K. N., & Eltawil, A. M. (2023) · Proceedings of the IEEE · doi:10.1109/jproc.2023.3268092 · Cited by 35
“Stuck-at-fault (SAF) is a common problem of the ReRAM crossbar array that occurs due to temporary writing failures (TWFs) [96], [97].”
BackgroundSpiking Neuron Hardware-Level Fault Modeling
El-Sayed, S. A., Spyrou, T., Pavlidis, A., Afacan, E., Camuñas-Mesa, L. A., Linares-Barranco, B., & Stratigopoulos, H. (2020) · IEEE International On-Line Testing Symposium (IOLTS) · doi:10.1109/iolts50870.2020.9159745 · Cited by 30
“…experiments have been performed using behavioral-level faults at the synapse and neuron level [3], static and transient bit flips in data-paths and memories [4]–[6], and stuck-at faults at gate-level [7], at quantizing activations [8], or at the conductance of memristors in memristor crossbars [9].”
MethodEffect of Program Error in Memristive Neural Network With Weight Quantization
Kim, T., Kim, S., Hong, K., Park, J., Youn, S., Lee, J., Park, B., & Kim, H. (2022) · IEEE Transactions on Electron Devices · doi:10.1109/ted.2022.3169112 · Cited by 31
Stuck-at-Fault Immunity Enhancement of Memristor-Based Edge AI Systems
Oli-Uz-Zaman, M., Khan, S. A., Oswald, W., Liao, Z., & Wang, J. (2022) · IEEE Journal on Emerging and Selected Topics in Circuits and Systems · doi:10.1109/jetcas.2022.3207687 · Cited by 32
“Previously proposed hardware-based solutions in [20], [21], [22], [23], and [24] have some limitations.”
BackgroundBuild reliable and efficient neuromorphic design with memristor technology
Li, B., Yan, B., Liu, C., & Li, H. H. (2019) · Asia and South Pacific Design Automation Conference (ASP-DAC) · doi:10.1145/3287624.3288744 · Cited by 29
“For example, the remapping in [27, 29] and the retraining approaches in [8] considered the distribution of weights of neural network models and the distribution of device defects.”
BackgroundMethodDesign of fault-tolerant neuromorphic computing systems
Liu, M., Xia, L., Wang, Y., & Chakrabarty, K. (2018) · IEEE European Test Symposium (ETS) · doi:10.1109/ets.2018.8400693 · Cited by 28
“A reconfigurable redundancy scheme was proposed to tolerate dynamic faults [40].”
Self-citationPrecise and scalable analogue matrix equation solving using resistive random-access memory chips
Zuo, P., Wang, Q., Luo, Y., Xie, R. H., Wang, S., Cheng, Z., Bao, L., Wang, Z., Cai, Y., Huang, R., et al. (2025) · Nature Electronics · doi:10.1038/s41928-025-01477-0 · Cited by 43
“If defects like stuck-at faults are detected, established methods, such as redundant with parallel arrays 60 , could be applied.”
Pruning of Deep Neural Networks for Fault-Tolerant Memristor-based Accelerators
Chen, C.-Y., & Chakrabarty, K. (2021) · ACM/IEEE Design Automation Conference (DAC) · doi:10.1109/dac18074.2021.9586269 · Cited by 24
Self-citationIntrinsic variation effect in memristive neural network with weight quantization
Park, J., Song, M. S., Youn, S., Kim, T., Kim, S., Hong, K., & Kim, H. (2022) · Nanotechnology · doi:10.1088/1361-6528/ac7651 · Cited by 23
Layer ensemble averaging for fault tolerance in memristive neural networks
Yousuf, O., Hoskins, B. D., Ramu, K., Fream, M., Borders, W. A., Madhavan, A., Daniels, M. W., Dienstfrey, A., McClelland, J. J., Lueker-Boden, M., et al. (2025) · Nature Communications · doi:10.1038/s41467-025-56319-6 · Cited by 34
“We offer comparisons with two existing state-of-the-art hardware-oriented fault tolerance schemes Mapping Algorithm with inner fault tolerant ability (MAO) 27,33 and the Committee Machines (CM) algorithm 26 .”
HDC-IM: Hyperdimensional Computing In-Memory Architecture based on RRAM
Liu, J., Ma, M., Zhu, Z., Wang, Y., & Yang, H. (2019) · IEEE International Conference on Electronics, Circuits and Systems (ICECS) · doi:10.1109/icecs46596.2019.8964906 · Cited by 22
“In this case, the accuracy of the general NN algorithm will be greatly reduced, if the device’s SAFs reach 20%, the error rate of CNN on the MNIST dataset will rise by more than 80% [13].”
MethodSelf-citationCost- and Dataset-free Stuck-at Fault Mitigation for ReRAM-based Deep Learning Accelerators
Jung, G., Fouda, M. E., Lee, S., Lee, J., Eltawil, A. M., & Kurdahi, F. J. (2021) · Design, Automation, and Test in Europe · doi:10.23919/date51398.2021.9474226 · Cited by 25
FTT-NAS: Discovering Fault-tolerant Convolutional Neural Architecture
Ning, X., Ge, G., Li, W., Zhu, Z., Zheng, Y., Chen, X., Gao, Z., Wang, Y., & Yang, H. (2021) · ACM Transactions on Design Automation of Electronic Systems · doi:10.1145/3460288 · Cited by 20
“Besides the soft errors, common hard errors such as SAFs, caused by fabrication defects or limited endurance, could result in severe performance degradation [49].”
BackgroundSelf-citationFault-free: A Fault-resilient Deep Neural Network Accelerator based on Realistic ReRAM Devices
Shin, H., Kang, M., & Kim, L. (2021) · ACM/IEEE Design Automation Conference (DAC) · doi:10.1109/dac18074.2021.9586286 · Cited by 20
Rescuing RRAM-Based Computing From Static and Dynamic Faults
Lin, J., Wen, C.-D., Hu, X., Tang, T., Lin, I.-C., Wang, Y., & Xie, Y. (2020) · IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems · doi:10.1109/tcad.2020.3037316 · Cited by 21
Self-citationFault tolerance in memristive crossbar-based neuromorphic computing systems
Xu, Q., Chen, S., Geng, H., Yuan, B., Yu, B., Wu, F., & Huang, Z. (2019) · Integration · doi:10.1016/j.vlsi.2019.09.008 · Cited by 19
“Sometimes the hardware-level fault tolerant scheme does not allow changes to the well trained model [17].”
BackgroundReview of memristor based neuromorphic computation: opportunities, challenges and applications
Shekinah Archita, S., & Ravi, V. K. K. (2024) · Engineering Research Express · doi:10.1088/2631-8695/ad6662 · Cited by 25
Mapping Transformation Enabled High-Performance and Low-Energy Memristor-Based DNNs
Oli-Uz-Zaman, M., Khan, S. A., Yuan, G., Liao, Z., Fu, J., Ding, C., Wang, Y., & Wang, J. (2022) · Journal of Low Power Electronics and Applications · doi:10.3390/jlpea12010010 · Cited by 18
“Items [33] [34] [35] [21] [20] [24] [23] [22] [36] [25] This Work”
MethodSelf-citationImplementation of Neuronal Intrinsic Plasticity by Oscillatory Device in Spiking Neural Network
Wu, L., Wang, Z., Bao, L., Yu, Z., Chen, Q., Ling, Y., Qin, Y., Bao, S., Chen, Z., Bai, G., et al. (2022) · IEEE Transactions on Electron Devices · doi:10.1109/ted.2022.3152468 · Cited by 18
Shin, H., Kang, M., & Kim, L. (2022) · IEEE Transactions on Computers · doi:10.1109/tc.2022.3227871 · Cited by 16
“[14], [15], [18] propose solutions to the SAF problem without the retraining process.”
BackgroundMethodResultInfluentialAlgorithmic Fault Detection for RRAM-based Matrix Operations
Liu, M., Xia, L., Wang, Y., & Chakrabarty, K. (2020) · ACM Transactions on Design Automation of Electronic Systems · doi:10.1145/3386360 · Cited by 17
“Note that static faults can be easily detected by an off-line test method [10] and tolerated by redundant designs or re-mapping methods [51].”
MethodSelf-citationAnalyzing Fault Tolerance Behaviour in Memristor-based Crossbar for Neuromorphic Applications
Yadav, D. N., Datta, K., & Sengupta, I. (2020) · IEEE International Test Conference India (ITC India) · doi:10.1109/itcindia49857.2020.9171788 · Cited by 16
“…repeatedly to reduce error; however, the method needs backtracking if the current step results in a worse solution. iii) Redundant Neurons : In [16], an approach has been proposed where additional memristors are used to handle the performance degradation caused by faulty memris-tors by making…”
BackgroundMethodAll-Passive Hardware Implementation of Multilayer Perceptron Classifiers
Ananthakrishnan, A., & Allen, M. G. (2020) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2020.3016901 · Cited by 18
“1) Retraining: Having identified defect locations, soft SMLPs were trained again to evolve a new set of defect-aware optimal weights for the network [30], [31].”
BackgroundMethodOnline Fault Detection in ReRAM-Based Computing Systems for Inferencing
Liu, M., & Chakrabarty, K. (2022) · IEEE Transactions on Very Large Scale Integration (VLSI) Systems · doi:10.1109/tvlsi.2021.3139530 · Cited by 20
Self-citationTraining-Free Stuck-At Fault Mitigation for ReRAM-Based Deep Learning Accelerators
Quan, C., Fouda, M. E., Lee, S., Jung, G., Lee, J., Eltawil, A. M., & Kurdahi, F. J. (2022) · IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems · doi:10.1109/tcad.2022.3222288 · Cited by 15
“In particular, nonidealities in RCAs, such as stuck-at fault (SAF) [7], [8], IR drop [9], [10], [11], [12], and device variabilities [13], [14], have been shown to degrade the quality of application results severely.”
BackgroundLi, W., Read, J., Jiang, H., & Yu, S. (2022) · IEEE Journal on Emerging and Selected Topics in Circuits and Systems · doi:10.1109/jetcas.2022.3223031 · Cited by 18
“Furthermore, stuck-at faults may be addressed through physical transformation of the weight array [29] or allocation of redundant columns [30].”
BackgroundLong Live TIME: Improving Lifetime and Security for NVM-Based Training-in-Memory Systems
Cai, Y., Lin, Y., Xia, L., Chen, X., Han, S., Wang, Y., & Yang, H. (2020) · IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems · doi:10.1109/tcad.2020.2977079 · Cited by 14
“The approach proposed in [22] makes full use of all available mapping space to tolerant stuck-at faults of the”
BackgroundSelf-citationNguyen, T.-H., Imran, M., Choi, J., & Yang, J.-S. (2021) · ACM/IEEE Design Automation Conference (DAC) · doi:10.1109/dac18074.2021.9586112 · Cited by 14
ATT: A Fault-Tolerant ReRAM Accelerator for Attention-based Neural Networks
Guo, H., Peng, L., Zhang, J., Chen, Q., & LeCompte, T. (2020) · IEEE International Conference on Computer Design: VLSI in Computers and Processors (ICCD) · doi:10.1109/iccd50377.2020.00047 · Cited by 19
“To verify the efficiency of the proposed algorithm, we use two schemes as the reference: Ideal and RX [14].”
BackgroundMethodInfluentialEfficient Identification of Critical Faults in Memristor-Based Inferencing Accelerators
Chen, C.-Y., & Chakrabarty, K. (2021) · IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems · doi:10.1109/tcad.2021.3102894 · Cited by 15
Self-citationFault-Tolerant Neuromorphic Computing Systems
Chaudhuri, A., Liu, M., & Chakrabarty, K. (2019) · IEEE International Test Conference (ITC) · doi:10.1109/itc44170.2019.9000146 · Cited by 18
“A reconfigurable redundancy scheme has also been proposed to tolerate dynamic faults [71].”
BackgroundSelf-citationAccelerating AI Applications using Analog In-Memory Computing: Challenges and Opportunities
Channamadhavuni, S., Thijssen, S., Jha, S. K., & Ewetz, R. (2021) · ACM Great Lakes Symposium on VLSI (GLSVLSI) · doi:10.1145/3453688.3461746 · Cited by 15
Dynamic Task Remapping for Reliable CNN Training on ReRAM Crossbars
Tung, C.-H., Joardar, B. K., Pande, P. P., Doppa, J. R., Li, H. H., & Chakrabarty, K. (2023) · Design, Automation, and Test in Europe · doi:10.23919/date56975.2023.10137238 · Cited by 10
“Permanent faults lead to poor accuracy for both CNN training and inferencing [17].”
BackgroundSelf-citationTolerating Defects in Low-Power Neural Network Accelerators Via Retraining-Free Weight Approximation
Hosseini, F. S., Meng, F., Yang, C., Wen, W., & Cammarota, R. (2021) · ACM Transactions on Embedded Computing Systems · doi:10.1145/3477016 · Cited by 10
“Another class of DNN-specific defect mitigation techniques rely on an extra level of redundancy in DNN accelerators to handle defects [19, 21, 47].”
BackgroundAn, J., Oh, S., Nguyen, T. V., & Min, K. (2022) · Micromachines · doi:10.3390/mi13020273 · Cited by 10
Huang, Y., He, Y., Wang, J., Yue, J., Zhang, L., Zou, K., Yang, H., & Liu, Y. (2022) · IEEE Transactions on Circuits & Systems II Express Briefs · doi:10.1109/tcsii.2022.3163177 · Cited by 10
Self-citationReliability Improvement in RRAM-based DNN for Edge Computing
Oli-Uz-Zaman, M., Khan, S. A., Yuan, G., Wang, Y., Liao, Z., Fu, J., Ding, C., & Wang, J. (2022) · IEEE International Symposium on Circuits and Systems proceedings · doi:10.1109/iscas48785.2022.9937260 · Cited by 10
“On-Line Fault Detection [12], stuck-at-fault tolerance in RRAM computing systems [13] and recovering stuck-at-fault defects using matrix transformation [14] were also proposed 978-1-6654-8485-5/22/$31.00 ©2022 IEEE 581 to restore the accuracy drop because of SAFs.”
BackgroundInfluentialSelf-citationEfficient grouping approach for fault tolerant weight mapping in memristive crossbar array
Yadav, D. N., Thangkhiew, P. L., Chakraborty, S., & Sengupta, I. (2023) · Memories - Materials Devices Circuits and Systems · doi:10.1016/j.memori.2023.100045 · Cited by 11
Hardware Fault Tolerance for Binary RRAM Crossbars
Chaudhuri, A., Yan, B., Chen, Y., & Chakrabarty, K. (2019) · IEEE International Test Conference (ITC) · doi:10.1109/itc44170.2019.9000179 · Cited by 11
“We next consider a random distribution (non-sparse or dense) of faulty devices, a scenario highlighted in [19].”
BackgroundMethodSelf-citationWESCO: Weight-encoded Reliability and Security Co-design for In-memory Computing Systems
Zhang, J., Wang, C., Cai, Y., Zhu, Z., Kline, D. W., Yang, H., & Wang, Y. (2022) · IEEE Computer Society Annual Symposium on VLSI · doi:10.1109/isvlsi54635.2022.00065 · Cited by 13
“There are four main classes of error correction methods for existing DNN accelerators: First is redundant replacement, including row/column-level replacement [8] and Dual/Triple Modular Redundancy (DMR/TMR)) [9].”
MethodSelf-citationOn-Line Fault Protection for ReRAM-Based Neural Networks
Li, W., Wang, Y., Liu, C., He, Y., Liu, L., Li, H., & Li, X. (2022) · IEEE Transactions on Computers · doi:10.1109/tc.2022.3160345 · Cited by 10
“[17] proposes to remap the model weights with redundant memristor columns.”
BackgroundSelf-citationRRAM Crossbar-Based Fault-Tolerant Binary Neural Networks (BNNs)
Gebregiorgis, A., Zografou, A., & Hamdioui, S. (2022) · IEEE European Test Symposium (ETS) · doi:10.1109/ets54262.2022.9810414 · Cited by 9
“Several software and hardware-based fault tolerance approaches have been proposed to mitigate the impact of SAF in RRAM-based CIM [4], [5], [6], [7], [8], [9].”
MethodSEC-CiM: Selective Error Compensation for ReRAM-based Compute-in-Memory*
Bommana, A. R., Firouzi, F., Ogbogu, C., Joardar, B. K., Doppa, J. R., Pande, P. P., & Chakrabarty, K. (2024) · IEEE International Test Conference (ITC) · doi:10.1109/itc51657.2024.00037 · Cited by 8
Self-citationJiang, P., Song, D., Huang, M., Yang, F., Wang, L., Liu, P., Miao, X., & Wang, X. (2025) · Science China Information Sciences · doi:10.1007/s11432-024-4240-x · Cited by 10
“Both of these faults can significantly impact the accuracy of NN [36].”
Dahiya, A., Kumar, S., Rani, S., & Mukherjee, S. (2025) · ACS Applied Electronic Materials · doi:10.1021/acsaelm.5c01025 · Cited by 9
Kwak, M., Lee, S., Kim, S., & Hwang, H. (2020) · IEEE Electron Device Letters · doi:10.1109/led.2020.3008936 · Cited by 9
“Researchers have studied that system performance can be recovered from stuck-at-LRS faults by retraining in software [16]–[18] and using redundant devices [15].”
BackgroundEnabling High-Quality Uncertainty Quantification in a PIM Designed for Bayesian Neural Network
Li, X., Wu, B., Sun, G., Zhang, Z., Yuan, Z., Wang, R., Huang, R., Niu, D., Zheng, H., Lu, Z., et al. (2022) · IEEE International Symposium on High-Performance Computer Architecture (HPCA) · doi:10.1109/hpca53966.2022.00080 · Cited by 13
A Fully Automated Platform for Evaluating ReRAM Crossbars
Pelke, R., Staudigl, F., Thomas, N., Bosbach, N., Hossein, M., Cubero-Cascante, J., Poehls, L. B., Leupers, R., & Joseph, J. M. (2024) · IEEE Latin American Test Symposium (LATS) · doi:10.1109/lats62223.2024.10534593 · Cited by 8
An Empirical Fault Vulnerability Exploration of ReRAM-Based Process-in-Memory CNN Accelerators
Dorostkar, A., Farbeh, H., & Zarandi, H. R. (2024) · IEEE Transactions on Reliability · doi:10.1109/tr.2024.3405825 · Cited by 9
“Read disturbances in ReRAM memory are considered astransientfaults[36],[37].”
FARe: Fault-Aware GNN Training on ReRAM-Based PIM Accelerators
Dhingra, P., Ogbogu, C., Joardar, B. K., Doppa, J. R., Kalyanaraman, A., & Pande, P. P. (2024) · Design, Automation, and Test in Europe · doi:10.23919/date58400.2024.10546762 · Cited by 10
Wang, J., Xu, Q., Yuan, B., Chen, S., Yu, B., & Wu, F. (2020) · IEEE International Symposium on Circuits and Systems proceedings · doi:10.1109/iscas45731.2020.9180923 · Cited by 7
“[7] presented a re-mapping scheme with inner fault tolerance to eliminate the impact of SAFs.”
BackgroundDing, Y., An, J., Shen, J., Jia, S., Guo, J., Wang, L., Gong, T., Jiang, P., Wang, Y., Chen, Y., et al. (2022) · Advanced Electronic Materials · doi:10.1002/aelm.202200870 · Cited by 7
Self-citationNguyen, T.-H., Imran, M., Choi, J., & Yang, J.-S. (2023) · IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems · doi:10.1109/tcad.2023.3240659 · Cited by 9
“Existingworks on tolerating stuck-at faults in emerging NVMs have aimed for neuromorphic applications [19], [20], [21], [22], [23].”
BackgroundMethodThe Impact of Device Uniformity on Functionality of Analog Passively-Integrated Memristive Circuits
Fahimi, Z., Mahmoodi, M. R., Klachko, M., Nili, H., & Strukov, D. B. (2021) · IEEE Transactions on Circuits and Systems I Regular Papers · doi:10.1109/tcsi.2021.3097282 · Cited by 13
Towards Resilient Deployment of In-Memory Neural Networks with High Throughput
Zhang, B., & Ewetz, R. (2021) · ACM/IEEE Design Automation Conference (DAC) · doi:10.1109/dac18074.2021.9586185 · Cited by 7
Fast and Low-Cost Mitigation of ReRAM Variability for Deep Learning Applications
Lee, S., Fouda, M. E., Lee, J., Eltawil, A. M., & Kurdahi, F. J. (2021) · IEEE International Conference on Computer Design: VLSI in Computers and Processors (ICCD) · doi:10.1109/iccd53106.2021.00051 · Cited by 7
FTT-NAS: Discovering Fault-Tolerant Neural Architecture.Preprint
Ning, X., Ge, G., Li, W., Zhu, Z., Zheng, Y., Chen, X., Gao, Z., Wang, Y., & Yang, H. (2020) · arXiv (Cornell University) · Cited by 9
Self-citationStuck-at Faults in ReRAM Neuromorphic Circuit Array and their Correction through Machine Learning
Sawal, V., & Wong, H. Y. (2024) · IEEE Latin American Electron Devices Conference (LAEDC) · doi:10.1109/laedc61552.2024.10555838 · Cited by 8
“Moreover, it is inevitable to have stuck-at faults in ReRAM after fabrications [17][18].”
BackgroundMalhotra, A., Wang, C., & Gupta, S. K. (2024) · Asia and South Pacific Design Automation Conference (ASP-DAC) · doi:10.1109/asp-dac58780.2024.10473947 · Cited by 6
“However, most solutions are incompatible with CiM [11], require additional DNN training [12], and/or have a large resource overhead [13].”
BackgroundOnline Fault Detection in ReRAM-Based Computing Systems by Monitoring Dynamic Power Consumption
Liu, M., & Chakrabarty, K. (2020) · IEEE International Test Conference (ITC) · doi:10.1109/itc44778.2020.9325259 · Cited by 6
“Various error-recovery approaches have been proposed for ReRAM crossbars, e.g., replacing faulty ReRAM crossbars with redundant crossbars and independent redundant columns [20], remapping weights to a ReRAM crossbar based on the spatial distribution of faults in the crossbar [11], [21], etc.”
MethodSelf-citationImproving Reliability of ReRAM-Based DNN Implementation through Novel Weight Distribution
Li, J., Mao, M., & Chakrabarti, C. (2019) · SiPS ... design and implementation - IEEE Workshop on Signal Processing Systems · doi:10.1109/sips47522.2019.9020318 · Cited by 5
“Among all fault types, stuck-at fault is the most dominant fault affecting the computation accuracy [21], so we only consider stuck-at faults for error injection.”
Reliability-Aware Ratioed Logic Operations for Energy-Efficient Computational ReRAM
Fernández, C., & Vourkas, I. (2022) · IFIP/IEEE International Conference on Very Large Scale Integration · doi:10.1109/vlsi-soc54400.2022.9939627 · Cited by 5
“Consequently, the RESET (SET) pulse will not be applied if the device is already in HRS (LRS), since that could lead to stuck-at-ON/OFF faults which mean permanent damage of the cells [19].”
BackgroundFault-Tolerant Neuromorphic Computing on Nanoscale Crossbar Architectures
Roth, R. M. (2022) · IEEE Information Theory Workshop (ITW) · doi:10.1109/itw54588.2022.9965851 · Cited by 9
“Several recent papers have suggested to address these errors through hardware testing and sparing mechanisms; see [21],[28],[44],[45].”
BackgroundMa, Y., Zheng, L., & Zhou, P. (2023) · ACM Journal on Emerging Technologies in Computing Systems · doi:10.1145/3585518 · Cited by 6
Online Fault-Tolerance for Memristive Neuromorphic Fabric Based on Local Approximation
Ahmed, S. T., Rakhmatullin, R., & Tahoori, M. B. (2023) · IEEE European Test Symposium (ETS) · doi:10.1109/ets56758.2023.10174237 · Cited by 5
“developed an algorithm for mapping weight matrices to RRAM conductance that uses redundancy [13].”
MethodRIBoNN: Designing Robust In-Memory Binary Neural Network Accelerators
Kundu, S., Malhotra, A., Raha, A., Gupta, S. K., & Basu, K. (2022) · IEEE International Test Conference (ITC) · doi:10.1109/itc50671.2022.00061 · Cited by 5
“The permanent open switch defects in the RRAM array leads to SA1 behavior, and SA0 faults can be caused by the OverForming (OF) defect or reset failure [4].”
BackgroundAguirre, F. L., Pazos, S. M., Palumbo, F., Morell, A., Suñé, J., & Miranda, E. A. (2021) · Electronics · doi:10.3390/electronics10192427 · Cited by 5
“Four additional works (Zhang et al. [55,56] (2019) and Xia et al. [22,26] (2017) and (2018)) discuss SAF mitigation techniques, but provide oversimplified CPA and memristor modelling approaches.”
BackgroundMethodResultInfluentialFault Diagnosis in ReCAM Arrays
Hezayyin, H. G., Mayahinia, M., & Tahoori, M. B. (2025) · IEEE International On-Line Testing Symposium (IOLTS) · doi:10.1109/iolts65288.2025.11117144 · Cited by 4
Wu, T., Ding, C., Zhou, W., Cheng, Y., Cheng, X., Wang, S., Xu, W., Shi, C., Liu, Z., & Wong, N. (2026) · ACM Transactions on Design Automation of Electronic Systems · doi:10.1145/3801559 · Cited by 9
Testing ReRAM-based TCAM for Computation-in-Memory Applications
Hezayyin, H. G., Mayahinia, M., & Tahoori, M. B. (2024) · doi:10.1109/dttis62212.2024.10780399 · Cited by 4
“In the stuck-at-on (stuck-at-open), regardless of the gate voltage, the transistor always (never) conducts the current [8].”
A Hierarchical Fault-Tolerant and Cost Effective Framework for RRAM Based Neural Computing Systems
Vatti, C. S., Rakesh, M. B., Reddy, P. R. T., & Acharyya, A. (2022) · IEEE Transactions on Circuits & Systems II Express Briefs · doi:10.1109/tcsii.2022.3144193 · Cited by 4
“RRAM writes are reduced by 97% which is better than approaches [2], [6], [9], [10], but marginally less than [11].”
BackgroundResultTOSA: Tolerating Stuck-At-Faults in Edge-based RRAM Inference Accelerators
Zhao, S., Liu, B., Wang, F., Feng, D., & Chen, X. (2023) · IEEE International Conference on Parallel and Distributed Systems (ICPADS) · doi:10.1109/icpads60453.2023.00172 · Cited by 5
“…correction modules aim to compensate for defects and achieve resilient com-1181 [8], redundancy [9]–[11], retraining [9], [11]–[14], remapping [10] Nonlinearity/ Asymmetry √ S programming pulses [15], hardware-assistance [16] Device-to-Device (D2D) Variation √ S retraining [9], [11]–[14]…”
BackgroundMethodKim, H., An, Y., Kim, M.-C., Heo, G.-C., & Shim, Y. S. (2024) · IEEE Transactions on Circuits & Systems II Express Briefs · doi:10.1109/tcsii.2024.3485178 · Cited by 5
“RTN introduces random fluctuations in read currents [7], SAFs cause devices to get stuck in HRS or LRS with a certain probability [8], and endurance failures degrade HRS values [9].”
FAMCroNA: Fault Analysis in Memristive Crossbars for Neuromorphic Applications
Yadav, D. N., Thangkhiew, P. L., Datta, K., Chakraborty, S., Drechsler, R., & Sengupta, I. (2022) · Journal of Electronic Testing · doi:10.1007/s10836-022-06001-2 · Cited by 5
“In the worst case, the process must be reverted if the solu-tion is undesirable from the previous steps. iii) Redundant Neurons : The authors in [36, 43] proposed a solution by using additional memristors in place of the faulty memristors to handle performance degradation.”
BackgroundMethodInfluentialMalhotra, A., Wang, C., & Gupta, S. K. (2023) · ACM/IEEE Design Automation Conference (DAC) · doi:10.1109/dac56929.2023.10247835 · Cited by 4
“The faults are distributed randomly and uniformly across each layer of each of the models, as in similar works [7].”
BackgroundMethodResultInfluentialTowards Functionally Robust AI Accelerators
Banerjee, S., Chen, C.-Y., Talukdar, J., Hung, S.-C., Chaudhuri, A., Nikdast, M., & Chakrabarty, K. (2021) · IEEE Microelectronics Design & Test Symposium (MDTS) · doi:10.1109/mdts52103.2021.9476140 · Cited by 4
“Fault detection [24], [25] and localization [26] can be followed by mitigation methods such as remapping and retraining to restore the classification accuracy [27]–[29].”
Self-citationMulti-Fidelity Nonideality Simulation and Evaluation Framework for Resistive Neuromorphic Computing
Quan, C., Fouda, M. E., Lee, S., & Lee, J. (2022) · Asilomar Conference on Signals, Systems and Computers (ACSSC) · doi:10.1109/ieeeconf56349.2022.10052098 · Cited by 3
“Here we briefly introduce commonly modeled nonidealities in RCAs including IR drop [7]–[10], stuck-at fault (SAF) [11]– [14], variability [15], [16], and read noise [16].”
BackgroundMethodInfluentialDouble Adjacent Error Correction in RRAM Matrix Multiplication using Weighted Checksums
Pinto, K. X., Kodali, K., Das, A., & Touba, N. A. (2024) · IEEE International On-Line Testing Symposium (IOLTS) · doi:10.1109/iolts60994.2024.10616083 · Cited by 3
“The primary challenge is RRAM’s low yield and error prone nature compared to existing technologies [3] [4].”
BackgroundZhang, B., Uysal, N., Fan, D., & Ewetz, R. (2020) · ACM Great Lakes Symposium on VLSI (GLSVLSI) · doi:10.1145/3386263.3406910 · Cited by 5
“First, the mem-ristors in the rows of the MCA connected to the neurons t and s are programmed to realize ¯ w s using the technique in [18].”
BackgroundMethodResultInfluential
Metadata
- Authors
- Xia, L., Huangfu, W., Tang, T., Yin, X., Chakrabarty, K., Xie, Y., Wang, Y., & Yang, H.
- Affiliations
- Tsinghua University (CN)
- University of California, Santa Barbara (US)
- Duke University (US)
- Venue
- IEEE Journal on Emerging and Selected Topics in Circuits and Systems
- Published
- 2018-03
- Volume
- 8
- Issue
- 1
- Pages
- 102-115
- ISSN
- 2156-3357 (ISSN-L)
- 2156-3365
- OpenAlex
- W2768104155
- Cited by
- 177
- References
- 31
- Citations by year
- 2026: 17 · 2025: 23 · 2024: 25 · 2023: 18 · 2022: 33 · 2021: 20 · 2020: 23 · 2019: 13 · 2018: 5
- FWCI
- 4.66
- Type
- journal article
- Language
- en
- Open access
- closed
- Licence
- Locations
- IEEE Journal on Emerging and Selected Topics in Circuits and Systems (publishedVersion)
- Rare & Special e-Zone (The Hong Kong University of Science and Technology) (submittedVersion)
- Funders
- National Key Research and Development Program of China (2017YFA0207600)
- National Natural Science Foundation of China (61373026, 61622403, 61621091)
- Joint Fund of Equipment Pre-Research and Ministry of Education (6141A02022608)
- U.S. Department of Energy, Office of Science, Office of Advanced Scientific Computing Research (DE-SC0013553)
- NSF (1461698, 1500848)
- Qualcomm
- Topics
- Physical Sciences › Engineering › Electrical and Electronic Engineering › Advanced Memory and Neural Computing
- Physical Sciences › Computer Science › Hardware and Architecture › VLSI and Analog Circuit Testing
- Physical Sciences › Engineering › Electrical and Electronic Engineering › Ferroelectric and Negative Capacitance Devices
- Keywords
- stuck-at faults, fault tolerance, resistive random access memory, redundancy schemes, in-memory computing, energy overhead, MNIST recognition
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.