Neuro-Inspired Computing With Emerging Nonvolatile Memorys
The DOI matches its Crossref record; 134 of the 141 references it cites were found in a registry or an index.
Abstract
This comprehensive review summarizes state of the art, challenges, and prospects of the neuro-inspired computing with emerging nonvolatile memory devices. First, we discuss the demand for developing neuro-inspired architecture beyond today's von-Neumann architecture. Second, we summarize the various approaches to designing the neuromorphic hardware (digital versus analog, spiking versus nonspiking, online training versus offline training) and discuss why emerging nonvolatile memory is attractive for implementing the synapses in the neural network. Then, we discuss the desired device characteristics of the synaptic devices (e.g., multilevel states, weight update nonlinearity/asymmetry, variation/noise), and survey a few representative material systems and device prototypes reported in the literature that show the analog conductance tuning. These candidates include phase change memory, resistive memory, ferroelectric memory, floating-gate transistors, etc. Next, we introduce the crossbar array architecture to accelerate the weighted sum and weight update operations that are commonly used in the neuro-inspired machine learning algorithms, and review the recent progresses of array-level experimental demonstrations for pattern recognition tasks. In addition, we discuss the peripheral neuron circuit design issues and present a device-circuit-algorithm codesign methodology to evaluate the impact of nonideal device effects on the system-level performance (e.g., learning accuracy). Finally, we give an outlook on the customization of the learning algorithms for efficient hardware implementation.
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
“To implement neural network computing hardware with a high learning accuracy, a low nonlinearity of synaptic weights with repetitive input pulses and a large on/off synapse conductance ratio are required 36 .”
From Enhanced analog synaptic behavior of SiNx/a-Si bilayer memristors through Ge implantation (2020) · cites this paper as [36]
In the citing paper
“The achieved learning accuracy is very comparable to the state-of-the-art results in the literature 36,46,49 .”
From Enhanced analog synaptic behavior of SiNx/a-Si bilayer memristors through Ge implantation (2020) · cites this paper as [36]
In the citing paper
“A promising solution to this problem is the use of memristor-based systems, which can drastically reduce the energy consumption of AI5,6, making it even conceivable to create self-powered edge AI systems that do not require batteries and can instead harvest energy from the environment.”
From Powering AI at the edge: A robust, memristor-based binarized neural network with near-memory computing and miniaturized solar cell (2024) · cites this paper as [5]
Citations
Cites
Deep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.90 · Cited by 235,474 · cited 1×
“For example, Microsoft’s Res-Net (which won the ImageNet 2015 image classification competition [5]) has more than 100 of layers [6].”
MethodVery Deep Convolutional Networks for Large-Scale Image Recognition
Simonyan, K., & Zisserman, A. (2014) · arXiv · doi:10.48550/arxiv.1409.1556 · Cited by 114,148 · cited 1×
“The adoption of low-precision weight is most suitable for the feedforward inference stage of the CNN models, including LeNet5 [127], AlexNet [3], and VGG [128] in an order of complexity.”
MethodImageNet classification with deep convolutional neural networks
Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2017) · Communications of the ACM · doi:10.1145/3065386 · Cited by 112,155 · cited 2×
“In particular, convolutional neural network (CNN) [3] and recurrent neural network (RNN) [4] algorithms and their variants have proved their efficacy in a wide range of image, video, speech, and biomedical applications.”
BackgroundMethodLeCun, Y., Bengio, Y., & Hinton, G. (2015) · Nature · doi:10.1038/nature14539 · Cited by 84,126
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×
“The adoption of low-precision weight is most suitable for the feedforward inference stage of the CNN models, including LeNet5 [127], AlexNet [3], and VGG [128] in an order of complexity.”
MethodMastering the game of Go with deep neural networks and tree search
Silver, D., Huang, A., Maddison, C. J., Guez, A., Sifre, L., van den Driessche, G., Schrittwieser, J., Antonoglou, I., Panneershelvam, V., Lanctot, M., et al. (2016) · Nature · doi:10.1038/nature16961 · Cited by 15,612 · cited 1×
“, Google’s AlphaGo [1]).”
Speech recognition with deep recurrent neural networks
Graves, A., Mohamed, A.-R., & Hinton, G. (2013) · 2013 IEEE International Conference on Acoustics, Speech and Signal Processing · doi:10.1109/icassp.2013.6638947 · Cited by 8,893 · cited 1×
“In particular, convolutional neural network (CNN) [3] and recurrent neural network (RNN) [4] algorithms and their variants have proved their efficacy in a wide range of image, video, speech, and biomedical applications.”
BackgroundRedox‐Based Resistive Switching Memories – Nanoionic Mechanisms, Prospects, and Challenges
Waser, R., Dittmann, R., Staikov, G., & Szot, K. (2009) · Advanced Materials · doi:10.1002/adma.200900375 · Cited by 5,080 · cited 1×
“RRAM has two subcategories [44]: one is anion-based oxide random access memory (OxRAM) and the other one is cation-based conductive bridge random access memory (CBRAM) [45].”
BackgroundIn-Datacenter Performance Analysis of a Tensor Processing Unit
Jouppi, N. P., Young, C., Patil, N., Patterson, D., Agrawal, G., Bajwa, R., Bates, S., Bhatia, S., Boden, N., Borchers, A., et al. (2017) · Proceedings of the 44th Annual International Symposium on Computer Architecture · doi:10.1145/3079856.3080246 · Cited by 4,537
XNOR-Net: ImageNet Classification Using Binary Convolutional Neural Networks
Rastegari, M., Ordonez, V., Redmon, J., & Farhadi, A. (2016) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-46493-0_32 · Cited by 4,408 · cited 1×
“A similar work that constraints the weights and neurons to be (+1, −1), namely XNOR-Net [130], has been demonstrated for the classification of ImageNet data set with slight degradation of accuracy.”
BackgroundNanoscale Memristor Device as Synapse in Neuromorphic Systems
Jo, S. H., Chang, T., Ebong, I., Bhadviya, B. B., Mazumder, P., & Lu, W. (2010) · Nano Letters · doi:10.1021/nl904092h · Cited by 4,174 · cited 2×
“Representative analog synaptic device weight update behaviors from the literature: TiOx/TiO2 [80], PCMO [79], Ag:a-Si [77], AlOx/ HfO2 [76].”
BackgroundA million spiking-neuron integrated circuit with a scalable communication network and interface
Merolla, P. A., Arthur, J. V., Alvarez-Icaza, R., Cassidy, A. S., Sawada, J., Akopyan, F., Jackson, B. L., Imam, N., Guo, C., Nakamura, Y., et al. (2014) · Science · doi:10.1126/science.1254642 · Cited by 4,147 · cited 2×
“, Heidelbergs BrainScaleS [9], IBM’s TrueNorth [11], etc.”
MethodMemristive devices for computing
Yang, J. J., Strukov, D. B., & Stewart, D. R. (2013) · Nature Nanotechnology · doi:10.1038/nnano.2012.240 · Cited by 3,885 · cited 2×
“There are several comprehensive reviews on eNVMs for digital memory applications [20]–[24] and CMOS-based neuromorphic circuits [25]–[27].”
BackgroundEyeriss: An Energy-Efficient Reconfigurable Accelerator for Deep Convolutional Neural Networks
Chen, Y.-H., Krishna, T., Emer, J. S., & Sze, V. (2017) · IEEE Journal of Solid-State Circuits · doi:10.1109/jssc.2016.2616357 · Cited by 3,247 · cited 3×
“The estimates [41], based on the experimentally measured parameters of the memory cells, showed that the transfer to the 55-nm technology, with some improvements of auxiliary CMOS circuits, will allow the implementation of much larger networks with a similar performance lead over the most prospective digital networks…”
BackgroundMethodTraining 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 2×
“CNN accelerators [12] [13, 14, 15] Analog synapses: UCSB’s 12×12 crossbar array [37]”
BackgroundWong, 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×
“The resistance-based eNVMs include the spintransfer torque magnetic random access memory (STTMRAM) [43], phase change memory (PCM) [20], and resistive random access memory (RRAM) [21].”
BackgroundSelf-citationMemristors with diffusive dynamics as synaptic emulators for neuromorphic computing
Wang, Z., Joshi, S., Savel’ev, S. E., Jiang, H., Midya, R., Lin, P., Hu, M., Ge, N., Strachan, J. P., Li, Z., et al. (2017) · Nature Materials · doi:10.1038/nmat4756 · Cited by 2,358 · cited 1×
“…various devices, including Ag/Ag 2 S-based CBRAM [56], Cu/Cu 2 S-based CBRAM [57], Ag/GeS 2 -based CBRAM [58], Ag/Ge 30 Se 70 -based CBRAM [59], Ag/SiO x N y -based CBRAM [60], TiO x -based OxRAM [61], HfO x -based OxRAM [55], [62], [63], WO x -based OxRAM [64], [65], TaO x -based OxRAM [66], etc.”
BinaryConnect: Training Deep Neural Networks with binary weights during propagations
Courbariaux, M., Bengio, Y., & David, J.-P. (2015) · arXiv · doi:10.48550/arxiv.1511.00363 · Cited by 2,180 · cited 1×
“In the extreme case, the binary weight and neuron, namely Binary-Net [129], has been demonstrated for the classification of CIFAR data set with negligible degradation of accuracy.”
MethodNeuromorphic Silicon Neuron Circuits
Indiveri, G., Linares-Barranco, B., Hamilton, T. J., Schaik, A. V., Etienne-Cummings, R., Delbruck, T., Liu, S.-C., Dudek, P., Häfliger, P., Renaud, S., et al. (2011) · Frontiers in Neuroscience · doi:10.3389/fnins.2011.00073 · Cited by 1,821 · cited 1×
“There are several comprehensive reviews on eNVMs for digital memory applications [20]–[24] and CMOS-based neuromorphic circuits [25]–[27].”
BackgroundWong, H.-S. P., Raoux, S., Kim, S., Liang, J., Reifenberg, J. P., Rajendran, B., Asheghi, M., & Goodson, K. E. (2010) · Proceedings of the IEEE · doi:10.1109/jproc.2010.2070050 · Cited by 1,806 · cited 3×
“There are several comprehensive reviews on eNVMs for digital memory applications [20]–[24] and CMOS-based neuromorphic circuits [25]–[27].”
BackgroundShort-term plasticity and long-term potentiation mimicked in single inorganic synapses
Ohno, T., Hasegawa, T., Tsuruoka, T., Terabe, K., Gimzewski, J. K., & Aono, M. (2011) · Nature Materials · doi:10.1038/nmat3054 · Cited by 1,793 · cited 1×
“In addition to the analog conductance tuning capability, biologically realistic behaviors such as short-term memory, pair-pulse facilitation, and spike-timing-dependent plasticity have been emulated in various devices including Ag/Ag2S [54], Cu/Cu2S [55], Ag/GeS2 [56], Ag/Ge30Se70 [57] Ag/SiOxNy [58] based CBRAM, TiOx…”
BackgroundBuilding high-level features using large scale unsupervised learning
Le, Q. V. (2013) · 2013 IEEE International Conference on Acoustics, Speech and Signal Processing · doi:10.1109/icassp.2013.6639343 · Cited by 1,695 · cited 1×
“algorithm was able to successfully identify faces of cats from 10 million random images taken from YouTube videos [7].”
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 1×
“, PRIME [118], ISAAC [119], and Harmonica [120]) have been developed to support system-level design of neuromorphic accelerators, however they have limited considerations at the aforementioned nonideal device properties (i.”
BackgroundDeep Learning with Limited Numerical Precision
Gupta, S., Agrawal, A., Gopalakrishnan, K., & Narayanan, P. (2015) · arXiv · doi:10.48550/arxiv.1502.02551 · Cited by 1,364
Furber, S. B., Galluppi, F., Temple, S., & Plana, L. A. (2014) · Proceedings of the IEEE · doi:10.1109/jproc.2014.2304638 · Cited by 1,350 · cited 1×
“Specific designed accelerators such as Manchester’s SpiNNaker [8], Heidelberger’s BrainScaleS [9], and Google’s tensor processing unit (TPU) [10] have been developed to run large-scale neuromorphic and/or deep learning algorithms.”
MethodNanoelectronic programmable synapses based on phase change materials for brain-inspired computing
Kuzum, D., Jeyasingh, R. G. D., Lee, B., & Wong, H.-S. P. (2012) · Nano Letters · doi:10.1021/nl201040y · Cited by 1,267 · cited 1×
“The realization of PCM-based synaptic devices could be dated back to the pioneering work in [67], where the device conductance could be gradually increased or decreased with ~ 100 states by applying a sequence of programming pulses with increasing amplitudes.”
BackgroundSynaptic electronics: materials, devices and applications
Kuzum, D., Yu, S., & Philip Wong, H.-S. (2013) · Nanotechnology · doi:10.1088/0957-4484/24/38/382001 · Cited by 1,258 · cited 2×
“A special subset of eNVM devices that show multilevel resistance/conductance states could naturally emulate a synaptic device in the neural network, namely, resistive synaptic devices [17].”
BackgroundSelf-citationChi, 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×
“Recently, architectural simulator platforms (e.g., PRIME [118], ISAAC [119], and Harmonica [120]) have been developed to support system-level design of neuromorphic accelerators, however they have limited considerations at the aforementioned nonideal device properties (i.e., they only considered the weight precision a…”
Short-Term Memory to Long-Term Memory Transition in a Nanoscale Memristor
Chang, T., Jo, S.-H., & Lu, W. (2011) · ACS Nano · doi:10.1021/nn202983n · Cited by 1,058 · cited 1×
“In addition to the analog conductance tuning capability, biologically realistic behaviors such as short-term memory, pair-pulse facilitation, and spike-timing-dependent plasticity have been emulated in various devices, including Ag/Ag2S-based CBRAM [56], Cu/Cu2S-based CBRAM [57], Ag/GeS2-based CBRAM [58], Ag/Ge30Se70-…”
cuDNN: Efficient Primitives for Deep Learning
Chetlur, S., Woolley, C., Vandermersch, P., Cohen, J., Tran, J. J., Catanzaro, B., & Shelhamer, E. (2014) · arXiv (Cornell University) · doi:10.48550/arxiv.1410.0759 · Cited by 943 · cited 1×
“As off-the-shelf technologies, GPUs [32] or field-programmable gate arrays (FPGAs) [33] have been widely used for hardware acceleration for machine/deep learning.”
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
Kim, K.-H., Gaba, S., Wheeler, D., Cruz-Albrecht, J. M., Hussain, T., Srinivasa, N., & Lu, W. (2012) · Nano Letters · doi:10.1021/nl203687n · Cited by 924 · cited 1×
“[95] demonstrated onetime programming weights into 40 × 40 Ag: a-Si crossbar array.”
BackgroundElectrochemical metallization memories—fundamentals, applications, prospects
Valov, I., Waser, R., Jameson, J. R., & Kozicki, M. N. (2011) · Nanotechnology · doi:10.1088/0957-4484/22/25/254003 · Cited by 921 · cited 1×
“RRAM has two subcategories [44]: one is anion-based oxide random access memory (OxRAM) and the other one is cation-based conductive bridge random access memory (CBRAM) [45].”
BackgroundFace classification using electronic synapses
Yao, P., Wu, H., Gao, B., Eryilmaz, S. B., Huang, X., Zhang, W., Zhang, Q., Deng, N., Shi, L., Wong, H.-S. P., et al. (2017) · Nature Communications · doi:10.1038/ncomms15199 · Cited by 909 · cited 2×
“[39] demonstrated a one-layer perceptron for face recognition with 128 × 8 1 T1R analog RRAM array, as shown in the micrograph of the array.”
MethodBuilding high-level features using large scale unsupervised learning
Ranzato, M., Monga, R., Devin, M., Chen, K., Corrado, G. S., Dean, J., Le, Q. V., & Ng, A. Y. (2012) · International Conference on Machine Learning · Cited by 822
Yu, S., Wu, Y., Jeyasingh, R., Kuzum, D., & Wong, H.-S. P. (2011) · IEEE Transactions on Electron Devices · doi:10.1109/ted.2011.2147791 · Cited by 805
Self-citation14.5 Eyeriss: An energy-efficient reconfigurable accelerator for deep convolutional neural networks
Chen, Y.-H., Krishna, T., Emer, J., & Sze, V. (2016) · 2016 IEEE International Solid-State Circuits Conference (ISSCC) · doi:10.1109/isscc.2016.7418007 · Cited by 713
Hu, 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×
“[99] developed a 64 × 64 TiOx-based 1T1R array with 6-b (64 levels) offline training and a single-layer perceptron for MNIST image recognition.”
MethodA wafer-scale neuromorphic hardware system for large-scale neural modeling
Schemmel, J., Briiderle, D., Griibl, A., Hock, M., Meier, K., & Millner, S. (2010) · Proceedings of 2010 IEEE International Symposium on Circuits and Systems · doi:10.1109/iscas.2010.5536970 · Cited by 682 · cited 1×
“BrainScaleS platform is based on HICANN chip in 180 nm node that use analog neurons similar as the leaky integrate-and-fire model and digital synapses made of 4-bit 6-transistor SRAM cells and 4-bit digitalanalog-converter (DAC) to interface with analog neurons [35].”
MethodGovoreanu, B., Kar, G. S., Chen, Y.-Y., Paraschiv, V., Kubicek, S., Fantini, A., Radu, I. P., Goux, L., Clima, S., Degraeve, R., et al. (2011) · 2011 International Electron Devices Meeting · doi:10.1109/iedm.2011.6131652 · Cited by 624 · cited 1×
“10 nm without sacrificing their endurance, retention, and tuning accuracy, with some of the properties (such as the on/off conductance ratio) being actually improved [102].”
BackgroundSparse coding with memristor networks
Sheridan, P. M., Cai, F., Du, C., Ma, W., Zhang, Z., & Lu, W. D. (2017) · Nature Nanotechnology · doi:10.1038/nnano.2017.83 · Cited by 619 · cited 1×
“[38] demonstrated a 32×32 WOx crossbar array with offline trained analog weights for implementing the sparse coding algorithm for”
MethodKim, S., Du, C., Sheridan, P., Ma, W., Choi, S., & Lu, W. D. (2015) · Nano Letters · doi:10.1021/acs.nanolett.5b00697 · Cited by 603 · cited 1×
“In addition to the analog conductance tuning capability, biologically realistic behaviors such as short-term memory, pair-pulse facilitation, and spike-timing-dependent plasticity have been emulated in various devices, including Ag/Ag2S-based CBRAM [56], Cu/Cu2S-based CBRAM [57], Ag/GeS2-based CBRAM [58], Ag/Ge30Se70-…”
BackgroundFerroelectric FET analog synapse for acceleration of deep neural network training
Jerry, M., Chen, P.-Y., Zhang, J., Sharma, P., Ni, K., Yu, S., & Datta, S. (2017) · 2017 IEEE International Electron Devices Meeting (IEDM) · doi:10.1109/iedm.2017.8268338 · Cited by 599 · cited 2×
“from [88].”
MethodSelf-citationLarge-scale neuromorphic computing systems
Furber, S. (2016) · Journal of Neural Engineering · doi:10.1088/1741-2560/13/5/051001 · Cited by 567 · cited 1×
“There are several comprehensive reviews on eNVMs for digital memory applications [20]–[24] and CMOS-based neuromorphic circuits [25]–[27].”
BackgroundOn Spike-Timing-Dependent-Plasticity, Memristive Devices, and Building a Self-Learning Visual Cortex
Zamarreño-Ramos, C., Camuñas-Mesa, L. A., Pérez-Carrasco, J. A., Masquelier, T., Serrano-Gotarredona, T., & Linares-Barranco, B. (2011) · Frontiers in Neuroscience · doi:10.3389/fnins.2011.00026 · Cited by 558 · cited 1×
“One important biologically plausible learning rule is the spike-timing-dependent plasticity (STDP) [36].”
BackgroundHigh precision tuning of state for memristive devices by adaptable variation-tolerant algorithm
Alibart, F., Gao, L., Hoskins, B. D., & Strukov, D. B. (2012) · Nanotechnology · doi:10.1088/0957-4484/23/7/075201 · Cited by 557 · cited 3×
“4(d) [83].”
BackgroundEmerging Memory Technologies: Recent Trends and Prospects
Yu, S., & Chen, P.-Y. (2016) · IEEE Solid-State Circuits Magazine · doi:10.1109/mssc.2016.2546199 · Cited by 555 · cited 1×
“The eNVMs are mostly pursued as the next-generation storage-class memory technologies with aggressive industrial research and development [22].”
BackgroundSelf-citationYu, 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×
“Thus in the early design of RRAM based synapse, one-way RESET-only learning scheme was used in HfOx based synaptic device [71].”
MethodSelf-citationWoo, J., Moon, K., Song, J., Lee, S., Kwak, M., Park, J., & Hwang, H. (2016) · IEEE Electron Device Letters · doi:10.1109/led.2016.2582859 · Cited by 512 · cited 5×
“To make the SET process gradual, one way is the oxide stack engineering (i.e. bilayer oxides) to make weak or multiple weak filaments as demonstrated in TaOx/HfO2 [73] and AlOx/HfO2 [74] devices.”
BackgroundMethodInfluentialGokmen, T., & Vlasov, Y. (2016) · Frontiers in Neuroscience · doi:10.3389/fnins.2016.00333 · Cited by 491 · cited 2×
“In principle, the weight update can be performed in a fully parallel fashion on the entire array as the programming voltages can be applied from both ends (row and column) to the synaptic device [80], [93].”
BackgroundMoons, B., Uytterhoeven, R., Dehaene, W., & Verhelst, M. (2017) · 2017 IEEE International Solid-State Circuits Conference (ISSCC) · doi:10.1109/isscc.2017.7870353 · Cited by 479 · cited 3×
“To further improve the energy-efficiency, CMOS based ASIC accelerators [10] [12] [13, 14, 15] have been prototyped.”
BackgroundMethodDu, C., Ma, W., Chang, T., Sheridan, P., & Lu, W. D. (2015) · Advanced Functional Materials · doi:10.1002/adfm.201501427 · Cited by 444 · cited 1×
“In addition to the analog conductance tuning capability, biologically realistic behaviors such as short-term memory, pair-pulse facilitation, and spike-timing-dependent plasticity have been emulated in various devices, including Ag/Ag2S-based CBRAM [56], Cu/Cu2S-based CBRAM [57], Ag/GeS2-based CBRAM [58], Ag/Ge30Se70-…”
Chen, P.-Y., Peng, X., & Yu, S. (2017) · 2017 IEEE International Electron Devices Meeting (IEDM) · doi:10.1109/iedm.2017.8268337 · Cited by 435 · cited 1×
“Here we present a simulation platform NeuroSim [122] to evaluate system-level metrics such as learning accuracy, area, latency, and energy for online training with these realistic device properties.”
MethodSelf-citationUltrafast Synaptic Events in a Chalcogenide Memristor
Li, Y., Zhong, Y., Xu, L., Zhang, J., Xu, X., Sun, H., & Miao, X. (2013) · Scientific Reports · doi:10.1038/srep01619 · Cited by 423 · cited 1×
“Then various pulse programming schemes have been proposed by different groups to reduce the complexity and power consumption of neuromorphic circuits using the PCM devices [66, 67, 68, 69].”
MethodReflections on the memory wall
McKee, S. A. (2004) · Proceedings of the 1st conference on Computing frontiers · doi:10.1145/977091.977115 · Cited by 420 · cited 1×
“data movement between the microprocessor and off-chip memory/storage has become the bottleneck of the entire system [30].”
BackgroundMemristors for Energy‐Efficient New Computing Paradigms
Jeong, D. S., Kim, K. M., Kim, S., Choi, B. J., & Hwang, C. S. (2016) · Advanced Electronic Materials · doi:10.1002/aelm.201600090 · Cited by 413
Finding a roadmap to achieve large neuromorphic hardware systems
Hasler, J., & Marr, B. (2013) · Frontiers in Neuroscience · doi:10.3389/fnins.2013.00118 · Cited by 392 · cited 1×
“3 applications [20, 21, 22, 23, 24] and CMOS based neuromorphic circuits [25, 26, 27].”
BackgroundA 20nm 1.8V 8Gb PRAM with 40MB/s program bandwidth
Choi, Y., Song, I., Park, M.-H., Chung, H., Chang, S., Cho, B., Kim, J., Oh, Y., Kwon, D., Sunwoo, J., et al. (2012) · 2012 IEEE International Solid-State Circuits Conference · doi:10.1109/isscc.2012.6176872 · Cited by 316 · cited 1×
“For example, Samsung has reported an 8 Gb PCM prototype chip in 20 nm node featuring 40 MB/s write bandwidth [44].”
BackgroundSuri, M., Bichler, O., Querlioz, D., Cueto, O., Perniola, L., Sousa, V., Vuillaume, D., Gamrat, C., & DeSalvo, B. (2011) · 2011 International Electron Devices Meeting · doi:10.1109/iedm.2011.6131488 · Cited by 314 · cited 1×
“Then various pulse programming schemes have been proposed by different groups to reduce the complexity and power consumption of neuromorphic circuits using the PCM devices [66, 67, 68, 69].”
MethodAn 8 Mb Multi-Layered Cross-Point ReRAM Macro With 443 MB/s Write Throughput
Kawahara, A., Azuma, R., Ikeda, Y., Kawai, K., Katoh, Y., Hayakawa, Y., Tsuji, K., Yoneda, S., Himeno, A., Shimakawa, K., et al. (2013) · IEEE Journal of Solid-State Circuits · doi:10.1109/jssc.2012.2215121 · Cited by 310 · cited 1×
“Panasonic has commercial products of microcontrollers with megabyte-capacity embedded TaOx RRAM [48].”
BackgroundSeo, K., Kim, I., Jung, S., Jo, M., Park, S., Park, J., Shin, J., Biju, K. P., Kong, J., Lee, K., et al. (2011) · Nanotechnology · doi:10.1088/0957-4484/22/25/254023 · Cited by 288 · cited 1×
“…various devices, including Ag/Ag 2 S-based CBRAM [56], Cu/Cu 2 S-based CBRAM [57], Ag/GeS 2 -based CBRAM [58], Ag/Ge 30 Se 70 -based CBRAM [59], Ag/SiO x N y -based CBRAM [60], TiO x -based OxRAM [61], HfO x -based OxRAM [55], [62], [63], WO x -based OxRAM [64], [65], TaO x -based OxRAM [66], etc.”
Improving Analog Switching in HfO x -Based Resistive Memory With a Thermal Enhanced Layer
Wu, W., Wu, H., Gao, B., Deng, N., Yu, S., & Qian, H. (2017) · IEEE Electron Device Letters · doi:10.1109/led.2017.2719161 · Cited by 282 · cited 2×
“To make the SET process gradual, one way is the oxide stack engineering (i.e. bilayer oxides) to make weak or multiple weak filaments as demonstrated in TaOx/HfO2 [73] and AlOx/HfO2 [74] devices.”
BackgroundSelf-citationBackpropagation for energy-efficient neuromorphic computing
Esser, S. K., Appuswamy, R., Merolla, P. A., Arthur, J. V., & Modha, D. S. (2015) · neural information processing systems · Cited by 281
Oh, S., Kim, T., Kwak, M., Song, J., Woo, J., Jeon, S., Yoo, I. K., & Hwang, H. (2017) · IEEE Electron Device Letters · doi:10.1109/led.2017.2698083 · Cited by 277 · cited 1×
“the doped HfO2) to gradually tune the gate capacitance and consequently the threshold voltage (VT) and the channel conductance by the application of short voltage pulses to the gate [83, 84].”
BackgroundNovel ferroelectric FET based synapse for neuromorphic systems
Mulaosmanovic, H., Ocker, J., Muller, S., Noack, M., Muller, J., Polakowski, P., Mikolajick, T., & Slesazeck, S. (2017) · 2017 Symposium on VLSI Technology · doi:10.23919/vlsit.2017.7998165 · Cited by 276 · cited 1×
“, the doped HfO2) to gradually tune the gate capacitance and consequently the threshold voltage and the channel conductance by the application of short voltage pulses to the gate [86], [87].”
BackgroundMagnetoresistive Random Access Memory: The Path to Competitiveness and Scalability
Zhu, J.-G. (2008) · Proceedings of the IEEE · doi:10.1109/jproc.2008.2004313 · Cited by 260 · cited 1×
“The resistance-based eNVMs include the spintransfer torque magnetic random access memory (STTMRAM) [43], phase change memory (PCM) [20], and resistive random access memory (RRAM) [21].”
BackgroundAmbrogio, S., Balatti, S., Milo, V., Carboni, R., Wang, Z.-Q., Calderoni, A., Ramaswamy, N., & Ielmini, D. (2016) · IEEE Transactions on Electron Devices · doi:10.1109/ted.2016.2526647 · Cited by 256 · cited 6×
“L. Gao, et al. [97] proposed a scheme to implement the convolution kernel in CNN by unrolling 2D kernel matrix into 1D column vector and demonstrated the concept in a 12×12 HfOx crossbar array.”
BackgroundInfluentialUnsupervised Learning by Spike Timing Dependent Plasticity in Phase Change Memory (PCM) Synapses
Ambrogio, S., Ciocchini, N., Laudato, M., Milo, V., Pirovano, A., Fantini, P., & Ielmini, D. (2016) · Frontiers in Neuroscience · doi:10.3389/fnins.2016.00056 · Cited by 252 · cited 1×
“Then various pulse programming schemes have been proposed by different groups to reduce the complexity and power consumption of neuromorphic circuits using the PCM devices [68]–[71].”
XNOR-RRAM: A scalable and parallel resistive synaptic architecture for binary neural networks
Sun, X., Yin, S., Peng, X., Liu, R., Seo, J.-S., & Yu, S. (2018) · 2018 Design, Automation & Test in Europe Conference & Exhibition (DATE) · doi:10.23919/date.2018.8342235 · Cited by 246 · cited 1×
“With this principle, there are a few simulation works to explore the binary neural network with eNVMs [133]–[135].”
BackgroundSelf-citationDeep Learning on FPGAs: Past, Present, and FuturePreprint
Lacey, G., Taylor, G. W., & Areibi, S. M. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1602.04283 · Cited by 241 · cited 1×
“As off-the-shelf technologies, GPUs [32] or field-programmable gate arrays (FPGAs) [33] have been widely used for hardware acceleration for machine/deep learning.”
Method3-D Memristor Crossbars for Analog and Neuromorphic Computing Applications
Adam, G. C., Hoskins, B. D., Prezioso, M., Merrikh-Bayat, F., Chakrabarti, B., & Strukov, D. B. (2017) · IEEE Transactions on Electron Devices · doi:10.1109/ted.2016.2630925 · Cited by 234 · cited 2×
“Adapted from [106].”
Neuromorphic Silicon Neurons and Large-Scale Neural Networks: Challenges and Opportunities
Poon, C.-S., & Zhou, K. (2011) · Frontiers in Neuroscience · doi:10.3389/fnins.2011.00108 · Cited by 232
Mitigating effects of non-ideal synaptic device characteristics for on-chip learning
Chen, P.-Y., Lin, B., Wang, I.-T., Hou, T.-H., Ye, J., Vrudhula, S., Seo, J.-S., Cao, Y., & Yu, S. (2015) · 2015 IEEE/ACM International Conference on Computer-Aided Design (ICCAD) · doi:10.1109/iccad.2015.7372570 · Cited by 232 · cited 4×
“learning accuracy, as indicated by device-algorithm co-simulation studies [49, 51, 90, 115].”
BackgroundMethodResultInfluentialSelf-citationScaling-up resistive synaptic arrays for neuro-inspired architecture: Challenges and prospect
Yu, S., Chen, P.-Y., Cao, Y., Xia, L., Wang, Y., & Wu, H. (2015) · 2015 IEEE International Electron Devices Meeting (IEDM) · doi:10.1109/iedm.2015.7409718 · Cited by 219 · cited 2×
“Nevertheless, design challenges arise from the device level up to the architecture level when the crossbar array size is scaled up to solve practical problems [112].”
BackgroundMethodSelf-citationBinary neural network with 16 Mb RRAM macro chip for classification and online training
Yu, S., Li, Z., Chen, P.-Y., Wu, H., Gao, B., Wang, D., Wu, W., & Qian, H. (2016) · 2016 IEEE International Electron Devices Meeting (IEDM) · doi:10.1109/iedm.2016.7838429 · Cited by 219 · cited 3×
“If the multilevel states in the resistive synaptic devices are insufficient to meet the algorithmic precision requirement, there are two alternative solutions: First, multiple devices could be grouped to represent higher precision at the expensive of area and energy [42].”
BackgroundInfluentialSelf-citationYu, S., Gao, B., Fang, Z., Yu, H., Kang, J., & Wong, H.-S. P. (2012) · 2012 International Electron Devices Meeting · doi:10.1109/iedm.2012.6479018 · Cited by 213
Self-citationResistive memory device requirements for a neural algorithm accelerator
Agarwal, S., Plimpton, S. J., Hughart, D. R., Hsia, A. H., Richter, I., Cox, J. A., James, C. D., & Marinella, M. J. (2016) · 2016 International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn.2016.7727298 · Cited by 213 · cited 1×
“These nonideal behaviors commonly exist in today’s synaptic devices and are potentially harmful to the learning accuracy, as indicated by device-algorithm cosimulation studies [51], [53], [93], [117].”
BackgroundVisual Pattern Extraction Using Energy-Efficient “2-PCM Synapse” Neuromorphic Architecture
Bichler, O., Suri, M., Querlioz, D., Vuillaume, D., DeSalvo, B., & Gamrat, C. (2012) · IEEE Transactions on Electron Devices · doi:10.1109/ted.2012.2197951 · Cited by 212 · cited 1×
“[70]: one is used to implement synaptic potentiation (LTP-device) while the other one is used to implement synaptic depression (LTD-device).”
MethodBurr, 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 · cited 4×
“Adapted from [53].”
BackgroundMethodInfluentialNeuromorphic speech systems using advanced ReRAM-based synapse
Park, S., Sheri, A., Kim, J., Noh, J., Jang, J., Jeon, M., Lee, B., Lee, B. R., Lee, B. H., & Hwang, H. (2013) · 2013 IEEE International Electron Devices Meeting · doi:10.1109/iedm.2013.6724692 · Cited by 201 · cited 2×
“Representative analog synaptic device weight update behaviors from the literature: TiOx/TiO2 [80], PCMO [79], Ag:a-Si [77], AlOx/ HfO2 [76].”
BackgroundKim, S., Ishii, M., Lewis, S., Perri, T., BrightSky, M., Kim, W., Jordan, R., Burr, G. W., Sosa, N., Ray, A., et al. (2015) · 2015 IEEE International Electron Devices Meeting (IEDM) · doi:10.1109/iedm.2015.7409716 · Cited by 194 · cited 1×
“[40] demonstrated a 64-kb ( 256 × 256 ) PCM 1T1R array with on-chip leaky integration and fire neuron circuits for continuous in situ STDP learning.”
BackgroundGuo, X., Bayat, F. M., Bavandpour, M., Klachko, M., Mahmoodi, M. R., Prezioso, M., Likharev, K. K., & Strukov, D. B. (2017) · 2017 IEEE International Electron Devices Meeting (IEDM) · doi:10.1109/iedm.2017.8268341 · Cited by 192 · cited 1×
“The estimates [108], based on the experimentally measured parameters of the memory cells, showed that the transfer to the 55 nm technology, with some improvements of auxiliary CMOS circuits, will allow the implementation of much larger networks with a similar performance lead over the most”
MethodRRAM-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 · cited 1×
“On the other hand, the interfacial device typically shows both gradual SET and RESET processes, as demonstrated in Ag:a-Si [77], PrCaMnO3 (PCMO) [78], [79], and TaOx/TiO2 [80], [81] devices.”
BackgroundTowards artificial neurons and synapses: a materials point of view
Seok Jeong, D., Kim, I., Ziegler, M., & Kohlstedt, H. (2013) · RSC Advances · doi:10.1039/c2ra22507g · Cited by 189 · cited 1×
“There are also pioneering reviews on the synaptic devices which mostly focused on the material aspects of synaptic devices [17], [28], [29].”
BackgroundHardware realization of BSB recall function using memristor crossbar arrays
Hu, M., Li, H., Wu, Q., & Rose, G. S. (2012) · Proceedings of the 49th Annual Design Automation Conference · doi:10.1145/2228360.2228448 · Cited by 187 · cited 1×
“The resistive crossbar array architecture has been proposed for implementing the weighted sum (or matrix–vector multiplication, dot-product operation) [89], which is the mostly time-consuming step in the neuro-inspired learning algorithms.”
MethodWang, I.-T., Chang, C.-C., Chiu, L.-W., Chou, T., & Hou, T.-H. (2016) · Nanotechnology · doi:10.1088/0957-4484/27/36/365204 · Cited by 186 · cited 1×
“Figure 2 (a) shows an example of the TaOx/TiO2 device conductance under identical programming pulses [49, 50].”
BackgroundControlling the Synaptic Plasticity of a Cu2S Gap‐Type Atomic Switch
Nayak, A., Ohno, T., Tsuruoka, T., Terabe, K., Hasegawa, T., Gimzewski, J. K., & Aono, M. (2012) · Advanced Functional Materials · doi:10.1002/adfm.201200640 · Cited by 183 · cited 1×
“…and spike-timing-dependent plasticity have been emulated in various devices, including Ag/Ag 2 S-based CBRAM [56], Cu/Cu 2 S-based CBRAM [57], Ag/GeS 2 -based CBRAM [58], Ag/Ge 30 Se 70 -based CBRAM [59], Ag/SiO x N y -based CBRAM [60], TiO x -based OxRAM [61], HfO x -based OxRAM [55],…”
MethodElectronic system with memristive synapses for pattern recognition
Park, S., Chu, M., Kim, J., Noh, J., Jeon, M., Hun Lee, B., Hwang, H., Lee, B., & Lee, B.-G. (2015) · Scientific Reports · doi:10.1038/srep10123 · Cited by 181 · cited 1×
“[98] demonstrated a single-layer perceptron network in 32 × 6 PCMO crossbar array with off-chip neuron circuits on a printed circuit board.”
BackgroundReduction 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×
“The interconnect resistance effect can be mitigated by either relaxing the wire width [90] or remapping the weights of the weights in the algorithms to the conductance in the devices [91].”
BackgroundBinary convolutional neural network on RRAM
Tang, T., Xia, L., Li, B., Wang, Y., & Yang, H. (2017) · 2017 22nd Asia and South Pacific Design Automation Conference (ASP-DAC) · doi:10.1109/aspdac.2017.7858419 · Cited by 176 · cited 1×
“With this principle, there are a few simulation works to explore the binary neural network with eNVMs [133]–[135].”
BackgroundDesoli, G., Chawla, N., Boesch, T., Singh, S.-P., Guidetti, E., De Ambroggi, F., Majo, T., Zambotti, P., Ayodhyawasi, M., Singh, H., et al. (2017) · 2017 IEEE International Solid-State Circuits Conference (ISSCC) · doi:10.1109/isscc.2017.7870349 · Cited by 174 · cited 3×
“To further improve the energy-efficiency, CMOS based ASIC accelerators [10] [12] [13, 14, 15] have been prototyped.”
Burr, G. W., Narayanan, P., Shelby, R. M., Sidler, S., Boybat, I., di Nolfo, C., & Leblebici, Y. (2015) · 2015 IEEE International Electron Devices Meeting (IEDM) · doi:10.1109/iedm.2015.7409625 · Cited by 168
Nanoscale electronic synapses using phase change devices
Jackson, B. L., Rajendran, B., Corrado, G. S., Breitwisch, M., Burr, G. W., Cheek, R., Gopalakrishnan, K., Raoux, S., Rettner, C. T., Padilla, A., et al. (2013) · ACM Journal on Emerging Technologies in Computing Systems · doi:10.1145/2463585.2463588 · Cited by 163 · cited 1×
“Then various pulse programming schemes have been proposed by different groups to reduce the complexity and power consumption of neuro-morphic circuits using the PCM devices [68]–[71].”
Method3D-stackable crossbar resistive memory based on Field Assisted Superlinear Threshold (FAST) selector
Jo, S. H., Kumar, T., Narayanan, S., Lu, W. D., & Nazarian, H. (2014) · 2014 IEEE International Electron Devices Meeting · doi:10.1109/iedm.2014.7046999 · Cited by 162 · cited 1×
“, FAST selector [92]) may be needed to minimize the leakage current in other unselected rows/columns.”
BackgroundNbOx based oscillation neuron for neuromorphic computing
Gao, L., Chen, P.-Y., & Yu, S. (2017) · Applied Physics Letters · doi:10.1063/1.4991917 · Cited by 160 · cited 1×
“15(b) shows the measured threshold switching I-V characteristics of the Pt/NbO x /Pt device structure [115].”
BackgroundSelf-citationStochastic learning in oxide binary synaptic device for neuromorphic computing
Yu, S., Gao, B., Fang, Z., Yu, H. Y., Kang, J., & Wong, H. ‐. P. (2013) · Frontiers in Neuroscience · doi:10.3389/fnins.2013.00186 · Cited by 158 · cited 1×
“Second, recent work shows that binary synaptic devices with stochastic weight update may equivalently provide the properties of analog synapses for some simple neural networks [49], [50].”
BackgroundSelf-citationSpecifications of Nanoscale Devices and Circuits for Neuromorphic Computational Systems
Rajendran, B., Liu, Y., Seo, J.-S., Gopalakrishnan, K., Chang, L., Friedman, D. J., & Ritter, M. B. (2013) · IEEE Transactions on Electron Devices · doi:10.1109/ted.2012.2227969 · Cited by 158 · cited 1×
“based ones with SRAM synaptic arrays [19].”
Demonstration of Convolution Kernel Operation on Resistive Cross-Point Array
Gao, L., Chen, P.-Y., & Yu, S. (2016) · IEEE Electron Device Letters · doi:10.1109/led.2016.2573140 · Cited by 153 · cited 1×
“[100] proposed a scheme to implement the convolution kernel in CNN by unrolling 2-D kernel matrix into 1-D column vector and demonstrated the concept in a 12 × 12 HfOx crossbar array.”
MethodSelf-citationSuri, M., Bichler, O., Querlioz, D., Palma, G., Vianello, E., Vuillaume, D., Gamrat, C., & DeSalvo, B. (2012) · Technical digest - International Electron Devices Meeting/Technical digest · doi:10.1109/iedm.2012.6479017 · Cited by 149 · cited 2×
“The early design exploration of array-level performance is based on simulation only with the device model fitted with single device measurement data, as pioneered in [58] and [94].”
MethodFully parallel write/read in resistive synaptic array for accelerating on-chip learning
Gao, L., Wang, I., Chen, P.-Y., Vrudhula, S. B. K., Seo, J.-S., Cao, Y. K., Hou, T., & Yu, S. (2015) · Nanotechnology · doi:10.1088/0957-4484/26/45/455204 · Cited by 145 · cited 3×
“In principle, the weight update can be performed in a fully parallel fashion on the entire array as the programming voltages can be applied from both ends (row and column) to the synaptic device [80], [93].”
BackgroundSelf-citationSim, J., Park, J., Kim, M., Bae, D., Choi, Y., & Kim, L. (2016) · Digest of technical papers/Digest of technical papers - IEEE International Solid-State Circuits Conference · doi:10.1109/isscc.2016.7418008 · Cited by 142 · cited 3×
“To further improve the energy-efficiency, CMOS based ASIC accelerators [10] [12] [13, 14, 15] have been prototyped.”
BackgroundMethodAn 8Mb multi-layered cross-point ReRAM macro with 443MB/s write throughput
Kawahara, A., Azuma, R., Ikeda, Y., Kawai, K., Katoh, Y., Tanabe, K., Nakamura, T., Sumimoto, Y., Yamada, N., Nakai, N., et al. (2012) · Digest of technical papers/Digest of technical papers - IEEE International Solid-State Circuits Conference · doi:10.1109/isscc.2012.6177078 · Cited by 140 · cited 29×
“[41] S. Yu, Z. Li, P.-Y. Chen, H. Wu, B. Gao, D. Wang, W. Wu, H. Qian, "Binary neural network with 16 Mb RRAM macro chip for classification and online training," in IEEE International Electron Devices Meeting (IEDM), 2016.”
BackgroundInfluentialSchmitt, S., Klahn, J., Bellec, G., Grubl, A., Guttler, M., Hartel, A., Hartmann, S., Husmann, D., Husmann, K., Jeltsch, S., et al. (2017) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn.2017.7966125 · Cited by 138 · cited 8×
“Heidelberg’s BrainScaleS [9], IBM’s TrueNorth [11], etc.”
MethodInfluentialA 1 TOPS/W Analog Deep Machine-Learning Engine With Floating-Gate Storage in 0.13 µm CMOS
Lu, J., Young, S. R., Arel, I., & Holleman, J. H. (2014) · IEEE Journal of Solid-State Circuits · doi:10.1109/jssc.2014.2356197 · Cited by 126 · cited 1×
“[98] developed a machine learning prototype chip in 130 nm node with floating-gate synapses, exhibiting a remarkable energy efficiency 480 GOPS/W in the training mode and 1 TOPS/W in the”
BackgroundSynaptic Characteristics of Ag/AgInSbTe/Ta-Based Memristor for Pattern Recognition Applications
Zhang, Y., Li, Y., Wang, X., & Friedman, E. G. (2017) · IEEE Transactions on Electron Devices · doi:10.1109/ted.2017.2671433 · Cited by 126 · cited 1×
“In addition, the space-charge limited current in dielectrics may also show the multilevel conductance levels by charging and discharging the trapped states [82].”
BackgroundSchmitt, S., Klähn, J., Bellec, G., Grübl, A., Güttler, M., Hartel, A., Hartmann, S., Husmann, D., Husmann, K., Jeltsch, S., et al. (2017) · TUGraz OPEN Library (Graz University of Technology) · Cited by 125
Resistive Random Access Memory for Future Information Processing System
Wu, H., Wang, X. H., Gao, B., Deng, N., Lu, Z., Haukness, B., Bronner, G. B., & Qian, H. (2017) · Proceedings of the IEEE · doi:10.1109/jproc.2017.2684830 · Cited by 117 · cited 1×
“3 applications [20, 21, 22, 23, 24] and CMOS based neuromorphic circuits [25, 26, 27].”
BackgroundA multiply-add engine with monolithically integrated 3D memristor crossbar/CMOS hybrid circuit
Chakrabarti, B., Lastras-Montaño, M. Á., Adam, G. C., Prezioso, M., Hoskins, B. D., Payvand, M., Madhavan, A., Ghofrani, H. A., Theogarajan, L. S., Cheng, K.-T. T., et al. (2017) · Scientific Reports · doi:10.1038/srep42429 · Cited by 102 · cited 1×
“11 [106], [107].”
Programming Protocol Optimization for Analog Weight Tuning in Resistive Memories
Gao, L., Chen, P.-Y., & Yu, S. (2015) · IEEE Electron Device Letters · doi:10.1109/led.2015.2481819 · Cited by 101 · cited 2×
“Similar iterative programming schemes were demonstrated in HfO x devices as well [85].”
BackgroundMethodSelf-citation3D synaptic architecture with ultralow sub-10 fJ energy per spike for neuromorphic computation
Wang, I., Lin, Y.-C., Wang, Y., Hsu, C.-W., & Hou, T. (2014) · Technical digest - International Electron Devices Meeting/Technical digest · doi:10.1109/iedm.2014.7047127 · Cited by 101 · cited 1×
“On the other hand, the interfacial device typically shows both gradual SET and RESET processes, as demonstrated in Ag:a-Si [77], PrCaMnO3 (PCMO) [78], [79], and TaOx/TiO2 [80], [81] devices.”
BackgroundNeuro-inspired Computing Using Resistive Synaptic Devices
Yu, S. (2017) · Springer eBooks · doi:10.1007/978-3-319-54313-0 · Cited by 89 · cited 1×
“computing with low leakage power consumption [16].”
Self-citationTemperature-insensitive analog vector-by-matrix multiplier based on 55 nm NOR flash memory cells
Guo, X., Bayat, F. M., Prezioso, M., Chen, Y., Nguyen, B. Q., Do, N., & Strukov, D. B. (2017) · Custom Integrated Circuits Conference · doi:10.1109/cicc.2017.7993628 · Cited by 82 · cited 2×
“Custom design has been recently demonstrated for the industrial-grade 180-nm [108], [109] and 55-nm [110] [Fig.”
BackgroundGarbin, D., Bichler, O., Vianello, E., Rafhay, Q., Gamrat, C., Perniola, L., Ghibaudo, G., & DeSalvo, B. (2014) · Technical digest - International Electron Devices Meeting/Technical digest · doi:10.1109/iedm.2014.7047126 · Cited by 76 · cited 2×
“In addition to the analog conductance tuning capability, biologically realistic behaviors such as short-term memory, pair-pulse facilitation, and spike-timing-dependent plasticity have been emulated in various devices, including Ag/Ag2S-based CBRAM [56], Cu/Cu2S-based CBRAM [57], Ag/GeS2-based CBRAM [58], Ag/Ge30Se70-…”
ResultMNSIM: Simulation Platform for Memristor-based Neuromorphic Computing System
Xia, L., Li, B., Tang, T., Gu, P., Yin, X., Huangfu, W., Chen, P.-Y., Yu, S., Cao, Y. K., Wang, Y., et al. (2016) · Design, Automation, and Test in Europe · doi:10.3850/9783981537079_0549 · Cited by 76 · cited 1×
“On the other hand, MNSIM [121] is a circuit-level macromodel of neuro-inspired architecture, but the accuracy in this model is the output error of weighted sum (matrix–vector multiplication), which is just one step of the algorithm and thus it lacks the runtime learning accuracy of the entire algorithm.”
BackgroundSelf-citationChen, P.-Y., Kadetotad, D. V., Xu, Z., Mohanty, A., Lin, B., Ye, J., Vrudhula, S. B. K., Seo, J.-S., Cao, Y. K., & Yu, S. (2015) · Design, Automation, and Test in Europe · doi:10.7873/date.2015.0620 · Cited by 74 · cited 2×
“This is because the I-V nonlinearity of the resistive synaptic device may distort the weighted sum accuracy if using analog voltage, and it is also difficult to generate multiple bias levels within a small read voltage range from peripheral circuit designs perspective [90].”
BackgroundSelf-citationLiu, X., Mao, M., Liu, B., Li, B., Wang, Y., Jiang, H., Barnell, M. D., Wu, Q., Yang, J. J., Li, H. H., et al. (2016) · IEEE Transactions on Circuits and Systems I Regular Papers · doi:10.1109/tcsi.2016.2529279 · Cited by 70 · cited 1×
“, PRIME [118], ISAAC [119], and Harmonica [120]) have been developed to support system-level design of neuromorphic accelerators, however they have limited considerations at the aforementioned nonideal device properties (i.”
BackgroundDesign of Ternary Neural Network With 3-D Vertical RRAM Array
Li, Z., Chen, P.-Y., Xu, H., & Yu, S. (2017) · IEEE Transactions on Electron Devices · doi:10.1109/ted.2017.2697361 · Cited by 70 · cited 1×
“hybrid precision requirement for the weight propagation and weight update [127].”
BackgroundSelf-citationParallel Architecture With Resistive Crosspoint Array for Dictionary Learning Acceleration
Kadetotad, D. V., Xu, Z., Mohanty, A., Chen, P.-Y., Lin, B., Ye, J., Vrudhula, S. B. K., Yu, S., Cao, Y. K., & Seo, J.-S. (2015) · IEEE Journal on Emerging and Selected Topics in Circuits and Systems · doi:10.1109/jetcas.2015.2426495 · Cited by 63 · cited 2×
“14(a) shows an example of the silicon CMOS neuron design using this principle [113].”
MethodSelf-citationExperimental demonstration of array-level learning with phase change synaptic devices
Eryilmaz, S. B., Kuzum, D., Jeyasingh, R. G. D., Kim, S., BrightSky, M., Lam, C., & Wong, H.-S. P. (2013) · 2013 IEEE International Electron Devices Meeting · doi:10.1109/iedm.2013.6724691 · Cited by 51 · cited 1×
“[96] employed a Hopfield network consisting of a 10 × 10 PCM one-transistor-one-resistor (1T1R) array and ten recurrently connected software neurons for the implementation of associative learning.”
MethodPrezioso, M., Kataeva, I., Merrikh‐Bayat, F., Hoskins, B. D., Adam, G. C., Sota, T., Likharev, K. K., & Strukov, D. B. (2015) · Technical digest - International Electron Devices Meeting/Technical digest · doi:10.1109/iedm.2015.7409719 · Cited by 51 · cited 1×
“Al2O3/TiO2-x were recently reported in [37, 101].”
Chen, P.-Y., Kadetotad, D. V., Xu, Z., Mohanty, A., Lin, B., Ye, J., Vrudhula, S. B. K., Seo, J.-S., Cao, Y. K., & Yu, S. (2015) · Design, Automation, and Test in Europe · doi:10.5555/2755753.2755947 · Cited by 50
Self-citationAnalog-input analog-weight dot-product operation with Ag/a-Si/Pt memristive devices
Gao, L., Alibart, F., & Strukov, D. B. (2012) · IEEE/IFIP International Conference on VLSI and System-on-Chip (VLSI-SoC) · doi:10.1109/vlsi-soc.2012.7332082 · Cited by 48
High density neuromorphic system with Mo/Pr0.7Ca0.3MnO3 synapse and NbO2 IMT oscillator neuron
Moon, K., Cha, E., Park, J., Gi, S.-G., Chu, M., Baek, K., Lee, B., Oh, S., & Hwang, H. (2015) · Technical digest - International Electron Devices Meeting/Technical digest · doi:10.1109/iedm.2015.7409721 · Cited by 48 · cited 1×
“The NbO 2 -based device that exhibits meal–insulator– transition has been proposed as such oscillatory neuron [116].”
BackgroundRedesigning commercial floating-gate memory for analog computing applications
Bayat, F. M., Guo, X., Om'mani, H. A., Do, N., Likharev, K. K., & Strukov, D. B. (2015) · IEEE International Symposium on Circuits and Systems proceedings · doi:10.1109/iscas.2015.7169048 · Cited by 47 · cited 1×
“Custom design has been recently demonstrated for the industrial-grade 180-nm [108], [109] and 55-nm [110] [Fig.”
BackgroundCompact oscillation neuron exploiting metal-insulator-transition for neuromorphic computing
Chen, P.-Y., Seo, J.-S., Cao, Y. K., & Yu, S. (2016) · doi:10.1145/2966986.2967015 · Cited by 46
Self-citationMahalanabis, D., Barnaby, H., Gonzalez-Velo, Y., Kozicki, M. N., Vrudhula, S. B. K., & Dandamudi, P. (2014) · Solid-State Electronics · doi:10.1016/j.sse.2014.07.002 · Cited by 42 · cited 1×
“In addition to the analog conductance tuning capability, biologically realistic behaviors such as short-term memory, pair-pulse facilitation, and spike-timing-dependent plasticity have been emulated in various devices including Ag/Ag2S [54], Cu/Cu2S [55], Ag/GeS2 [56], Ag/Ge30Se70 [57] Ag/SiOxNy [58] based CBRAM, TiOx…”
BackgroundDesign Considerations of Selector Device in Cross-Point RRAM Array for Neuromorphic Computing
Woo, J., Peng, X., & Yu, S. (2018) · IEEE International Symposium on Circuits and Systems proceedings · doi:10.1109/iscas.2018.8351735 · Cited by 38 · cited 1×
“The exponential I-V selector, on the other hand, is less preferred in the neuro-inspired computing system as the exponential I-V of the selector introduces the distortion of the weighted sum accuracy [124].”
BackgroundSelf-citationTraining a Probabilistic Graphical Model With Resistive Switching Electronic Synapses
Eryilmaz, S. B., Neftci, E. O., Joshi, S., Kim, S., BrightSky, M. J., Lung, H.-L., Lam, C. H., Cauwenberghs, G., & Wong, H. ‐. P. (2016) · eScholarship (California Digital Library) · Cited by 38
A Deep Neural Network Compression Pipeline: Pruning, Quantization, Huffman EncodingPreprint
Han, S., Mao, H., & Dally, W. J. (2015) · arXiv (Cornell University) · Cited by 37
Model-based high-precision tuning of NOR flash memory cells for analog computing applications
Bayat, F. M., Guo, X., Klachko, M., Do, N., Likharev, K. K., & Strukov, D. B. (2016) · Conference digest · doi:10.1109/drc.2016.7548449 · Cited by 36 · cited 1×
“Custom design has been recently demonstrated for the industrial-grade 180-nm [108], [109] and 55-nm [110] [Fig.”
BackgroundTraining a Probabilistic Graphical Model With Resistive Switching Electronic Synapses
Eryilmaz, S. B., Neftci, E., Joshi, S., Kim, S., BrightSky, M., Lung, H.-L., Lam, C., Cauwenberghs, G., & Wong, H.-S. P. (2016) · IEEE Transactions on Electron Devices · doi:10.1109/ted.2016.2616483 · Cited by 34
Oxide-based analog synapse: Physical modeling, experimental characterization, and optimization
Gao, B., Wu, H., Kang, J., Yu, H., & Qian, H. (2016) · Technical digest - International Electron Devices Meeting/Technical digest · doi:10.1109/iedm.2016.7838367 · Cited by 27 · cited 1×
“The abrupt SET is attributed to the positive feedback between the filament growth speed and the electric field, resulting in the formation of a single dominate strong filament [74].”
BackgroundAn energy-efficient and high-throughput bitwise CNN on sneak-path-free digital ReRAM crossbar
Ni, L., Liu, Z., Song, W., Yang, J. J., Yu, H., Wang, K., & Wang, Y. (2017) · International Symposium on Low Power Electronics and Design (ISLPED) · doi:10.1109/islped.2017.8009177 · Cited by 25 · cited 1×
“With this principle, there are a few simulation works to explore the binary neural network with eNVMs [129, 130].”
BackgroundProgramming protocol optimization for analog weight tuning in resistive memories
Gao, L., & Yu, S. (2015) · Conference digest · doi:10.1109/drc.2015.7175619 · Cited by 7
Self-citation
Research works printed but not linked: 9
- jouppi, 2016, Google Supercharges Machine Learning Tasks with TPU Custom Chip
- manan, 2012, CBRAM devices as binary synapses for low-power stochastic neuromorphic systems: Auditory (cochlea) and visual (retina) cognitive processing applications, IEDM Tech Dig
- 0, The MNIST dataset of handwritten digits
- 0, Image
- liu, 2013, A 130.7 mm $^{2}~2$ -layer 32 Gb ReRAM memory device in 24nm technology, IEEE Int Solid-State Circuits Conf (ISSCC) Dig Tech Papers, 31.6.1
- 0, Theano
- govoreanu, 2011, $10\times 10$ nm2 Hf/HfOx crossbar resistive RAM with excellent performance, reliability and low-energy operation, IEDM Tech Dig
- 0, Yale Face Database
- 0, NeuroSim
Cited by
Memory devices and applications for in-memory computing
Sebastian, A., Le Gallo, M., Khaddam-Aljameh, R., & Eleftheriou, E. (2020) · Nature Nanotechnology · doi:10.1038/s41565-020-0655-z · Cited by 2,289
Fully hardware-implemented memristor convolutional neural network
Yao, P., Wu, H., Gao, B., Tang, J., Zhang, Q., Zhang, W., Yang, J. J., & Qian, H. (2020) · Nature · doi:10.1038/s41586-020-1942-4 · Cited by 2,179
Physics for neuromorphic computing
Marković, D., Mizrahi, A., Querlioz, D., & Grollier, J. (2020) · Nature Reviews Physics · doi:10.1038/s42254-020-0208-2 · Cited by 1,176
Resistive switching materials for information processing
Wang, Z., Wu, H., Burr, G. W., Hwang, C. S., Wang, K. L., Xia, Q., & Yang, J. J. (2020) · Nature Reviews Materials · doi:10.1038/s41578-019-0159-3 · Cited by 1,214
Two-dimensional materials for next-generation computing technologies
Liu, C., Chen, H., Wang, S., Liu, Q., Jiang, Y.-G., Zhang, D. W., Liu, M., & Zhou, P. (2020) · Nature Nanotechnology · doi:10.1038/s41565-020-0724-3 · Cited by 1,045
Metadata
- Authors
- Yu, S.
- Affiliations
- Arizona State University (US)
- Corresponding author
- Yu, Shimeng
- ORCID
- Yu, Shimeng
- Venue
- Proceedings of the IEEE
- Published
- 2018-02
- Volume
- 106
- Issue
- 2
- Pages
- 260-285
- ISSN
- 0018-9219 (ISSN-L)
- 1558-2256
- OpenAlex
- W2785141883
- Cited by
- 1,170
- References
- 141
- Citations by year
- 2026: 96 · 2025: 139 · 2024: 138 · 2023: 148 · 2022: 177 · 2021: 146 · 2020: 162 · 2019: 117 · 2018: 47
- FWCI
- 45.15
- Type
- journal article
- Language
- en
- Open access
- closed
- Licence
- Locations
- Proceedings of the IEEE (publishedVersion)
- Funders
- NSF-CCF (NSF-CCF-1552687)
- NSFCCF (NSFCCF-1715443)
- NSF-ECCS (NSF-ECCS-1701565)
- HDTRA (HDTRA1-17-1-0038)
- NSF/SRC (NSF/SRC E2CDA)
- ASCENT (one of the six SRC/DARPA JUMP Centers)
- Qualcomm
- Samsung
- Topics
- Physical Sciences › Engineering › Electrical and Electronic Engineering › Advanced Memory and Neural Computing
- Physical Sciences › Engineering › Electrical and Electronic Engineering › Ferroelectric and Negative Capacitance Devices
- Physical Sciences › Computer Science › Artificial Intelligence › Neural Networks and Reservoir Computing
- Keywords
- neuromorphic computing, neuromorphic hardware, non-volatile memory, artificial synapses, phase-change memory, resistive random access memory, ferroelectric memory, floating-gate transistors, pattern 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.