Training and operation of an integrated neuromorphic network based on metal-oxide memristors
Open accessThe DOI matches its Crossref record; 33 of the 33 references it cites were found in a registry or an index.
Abstract
Despite much progress in semiconductor integrated circuit technology, the extreme complexity of the human cerebral cortex, with its approximately 1014 synapses, makes the hardware implementation of neuromorphic networks with a comparable number of devices exceptionally challenging. To provide comparable complexity while operating much faster and with manageable power dissipation, networks based on circuits combining complementary metal-oxide-semiconductors (CMOSs) and adjustable two-terminal resistive devices (memristors) have been developed. In such circuits, the usual CMOS stack is augmented with one or several crossbar layers, with memristors at each crosspoint. There have recently been notable improvements in the fabrication of such memristive crossbars and their integration with CMOS circuits, including first demonstrations of their vertical integration. Separately, discrete memristors have been used as artificial synapses in neuromorphic networks. Very recently, such experiments have been extended to crossbar arrays of phase-change memristive devices. The adjustment of such devices, however, requires an additional transistor at each crosspoint, and hence these devices are much harder to scale than metal-oxide memristors, whose nonlinear current–voltage curves enable transistor-free operation. Here we report the experimental implementation of transistor-free metal-oxide memristor crossbars, with device variability sufficiently low to allow operation of integrated neural networks, in a simple network: a single-layer perceptron (an algorithm for linear classification). The network can be taught in situ using a coarse-grain variety of the delta rule algorithm to perform the perfect classification of 3 × 3-pixel black/white images into three classes (representing letters). This demonstration is an important step towards much larger and more complex memristive neuromorphic networks.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“In all these works the VMM operation required for the forward pass is performed by the memristor crossbar and the digitalized output vectors recorded by an acquisition printed circuit board. Then the output vector is processed by the training algorithm in software to determine how to update the synaptic weight after each training epoch.”
From Hardware implementation of memristor-based artificial neural networks (2024) · cites this paper as [105]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Crossbar arrays of two-terminal metal/insulator/metal (MIM) memristive devices can be fabricated easily using standard lithography and deposition techniques; this has been readily achieved by multiple groups57,64,68–70,85,93,105,216,217.”
From Hardware implementation of memristor-based artificial neural networks (2024) · cites this paper as [105]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“In the case where the memductance function G ( x, 0) is a smooth function x , then this memristor can be used as a synapse [3], [32], [35] in brainslike machines.”
From Everything You Wish to Know About Memristors But Are Afraid to Ask (2015) · cites this paper as [35]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“The progress in the most prospective, passive analog-grade ReRAM circuits13,17,21,22,24, however, has been much slower, mainly because of much stricter requirements for the uniformity of memory cells’ I–V characteristics (Supplementary Fig. 1).”
From 4K-memristor analog-grade passive crossbar circuit (2021) · cites this paper as [17]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“To reduce disturbance of already tuned half-selected devices in passively integrated crossbar circuits, the half-biasing scheme is adopted when applying write pulses17 (Supplementary Fig. 1).”
From 4K-memristor analog-grade passive crossbar circuit (2021) · cites this paper as [17]
Experimental — accuracy not yet measured
5 more claims about this paper
Create a free account to see every claim other papers make about this one.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
ImageNet 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 1×
“We believe that this demonstration is an important step toward the effective analog-hardware implementation of much more complex neuromorphic networks – first of all, multilayer-perceptron classifiers with deep learning 30 , and eventually also much more elaborate CrossNet-based cognitive systems.”
BackgroundIntroduction to the theory of neural computation
(1994) · Neural Networks · doi:10.1016/0893-6080(94)90014-0 · Cited by 6,401
Redox‐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
Nanoscale 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
Memristive 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 1×
“However, the practical implementation of such networks is still very challenging, due to the specific physical mechanism of resistance change in most prospective, metal-oxide based memristor – reversible modulation of the concentration profile of oxygen vacancies 11,20,21 .”
BackgroundSelf-citationWong, 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 2×
“However, the practical implementation of such networks is still very challenging, due to the specific physical mechanism of resistance change in most prospective, metal-oxide based memristor – reversible modulation of the concentration profile of oxygen vacancies 11,20,21 .”
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
Introduction to the Theory of Neural Computation
Hertz, J., Krogh, A., & Palmer, R. G. (2018) · CRC Press eBooks · doi:10.1201/9780429499661 · Cited by 1,772
LeCun, Y., Bottou, L., Orr, G., & Müller, K. (1998) · neural information processing systems · Cited by 1,333
LeCun, Y. A., Bottou, L., Orr, G. B., & Müller, K.-R. (2012) · Lecture Notes in Computer Science · doi:10.1007/978-3-642-35289-8_3 · Cited by 1,264
Chanthbouala, A., Garcia, V., Cherifi, R. O., Bouzehouane, K., Fusil, S., Moya, X., Xavier, S., Yamada, H., Deranlot, C., Mathur, N. D., et al. (2012) · Nature Materials · doi:10.1038/nmat3415 · Cited by 1,220
Experimental demonstration of associative memory with memristive neural networks
Pershin, Y. V., & Di Ventra, M. (2010) · Neural Networks · doi:10.1016/j.neunet.2010.05.001 · Cited by 1,041
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
Memristor−CMOS Hybrid Integrated Circuits for Reconfigurable Logic
Xia, Q., Robinett, W., Cumbie, M. W., Banerjee, N., Cardinali, T. J., Yang, J. J., Wu, W., Li, X., Tong, W. M., Strukov, D. B., et al. (2009) · Nano Letters · doi:10.1021/nl901874j · Cited by 671
Self-citationPattern classification by memristive crossbar circuits using ex situ and in situ training
Alibart, F., Zamanidoost, E., & Strukov, D. B. (2013) · Nature Communications · doi:10.1038/ncomms3072 · Cited by 642
Self-citationFinding 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
An 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
CARPENTER, M. B. (1976) · Cited by 293
Seo, 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
An Electronic Version of Pavlov's Dog
Ziegler, M., Soni, R., Patelczyk, T., Ignatov, M., Bartsch, T., Meuffels, P., & Kohlstedt, H. (2012) · Advanced Functional Materials · doi:10.1002/adfm.201200244 · Cited by 206
Brain-like associative learning using a nanoscale non-volatile phase change synaptic device array
Eryilmaz, S. B., Kuzum, D., Jeyasingh, R., Kim, S., BrightSky, M., Lam, C., & Wong, H.-S. P. (2014) · Frontiers in Neuroscience · doi:10.3389/fnins.2014.00205 · Cited by 205 · cited 1×
“19 Both these approaches are incompatible with the goal of reaching the extremely high density of neuromorphic networks, discussed above.”
BackgroundA 130.7-$\hbox{mm}^{2}$ 2-Layer 32-Gb ReRAM Memory Device in 24-nm Technology
Liu, T.-Y., Yan, T. H., Scheuerlein, R., Chen, Y., Lee, J. K., Balakrishnan, G., Yee, G., Zhang, H., Yap, A., Ouyang, J., et al. (2014) · IEEE Journal of Solid-State Circuits · doi:10.1109/jssc.2013.2280296 · Cited by 179
Kim, G. H., Lee, J. H., Ahn, Y., Jeon, W., Song, S. J., Seok, J. Y., Yoon, J. H., Yoon, K. J., Park, T. J., & Hwang, C. S. (2013) · Advanced Functional Materials · doi:10.1002/adfm.201202170 · Cited by 174
Four-dimensional address topology for circuits with stacked multilayer crossbar arrays
Strukov, D. B., & Williams, R. S. (2009) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.0906949106 · Cited by 153 · cited 2×
“The other operations, described by equations (1), (4) and (5), were performed by external electronics, but they are much less critical for network performance, and in future, larger CrossNets may be (at least partly) assisted by CMOS subsystems.”
BackgroundSelf-citationFerroelectric Artificial Synapses for Recognition of a Multishaded Image
Kaneko, Y., Nishitani, Y., & Ueda, M. (2014) · IEEE Transactions on Electron Devices · doi:10.1109/ted.2014.2331707 · Cited by 149
Optimization of the Backpropagation Algorithm for Training Multilayer Perceptrons
Schiffmann, W. H., Joost, M., & Werner, R. (1994) · Cited by 137
Hybrid CMOS/Nanoelectronic Circuits: Opportunities and Challenges
Likharev, K. K. (2008) · Journal of Nanoelectronics and Optoelectronics · doi:10.1166/jno.2008.301 · Cited by 132 · cited 2×
“al neural network using memristors integrated into a dense, transistor-free crossbar circuit. This crossbar performed, on the physical (Ohm’s-law) level the analog vector-by-matrix multiplication (2)-(3), which is by far the most computationally intensive part of operation of any neuromorphic network. The other operat…”
MethodInfluentialSelf-citationRedox-Based Resistive Switching Memories
Waser, R. M. (2012) · Journal of Nanoscience and Nanotechnology · doi:10.1166/jnn.2012.6652 · Cited by 116
CrossNets: Neuromorphic Hybrid CMOS/Nanoelectronic Networks
Likharev, K. K. (2011) · Science of Advanced Materials · doi:10.1166/sam.2011.1177 · Cited by 115 · cited 2×
“ficial neural network using memristors integrated into a dense, transistor-free crossbar circuit. This crossbar performed, on the physical (Ohm’s-law) level the analog vector-by-matrix multiplication (2)-(3), which is by far the most computationally intensive part of operation of any neuromorphic network. The other op…”
MethodInfluentialSelf-citationGovoreanu, B., Redolfi, A., Zhang, L., Adelmann, C., Popovici, M., Clima, S., Hody, H., Paraschiv, V., Radu, I. P., Franquet, A., et al. (2013) · 2013 IEEE International Electron Devices Meeting · doi:10.1109/iedm.2013.6724599 · Cited by 83 · cited 1×
“26,27 On the negative side, the scale makes the device-to-device reproducibility of device parameters, most importantly of the voltage required for memristor's electric forming 20,21 , hard to achieve with currently used fabrication technologies.”
BackgroundCross point arrays of 8 nm × 8 nm memristive devices fabricated with nanoimprint lithography
Pi, S., Lin, P., & Xia, Q. (2013) · Journal of Vacuum Science & Technology B, Nanotechnology and Microelectronics: Materials, Processing, Measurement, and Phenomena · doi:10.1116/1.4827021 · Cited by 81
Programmable cell array using rewritable solid-electrolyte switch integrated in 90nm CMOS
Miyamura, M., Nakaya, S., Tada, M., Sakamoto, T., Okamoto, K., Banno, N., Ishida, S., Ito, K., Hada, H., Sakimura, N., et al. (2011) · Digest of technical papers/Digest of technical papers - IEEE International Solid-State Circuits Conference · doi:10.1109/isscc.2011.5746296 · Cited by 43
Pattern classification by memristive crossbar circuits with ex-situ and in-situ training
Alibart, F., Zamanidoost, E., & Strukov, D. B. (2013) · HAL · Cited by 29
Self-citation
Research works printed but not linked: 1
- 6. Chevallier, C. J. et al. 0.13μm 64Mb multi-layered conductive metal-oxide memory. ISSCC'10 , 260-261 (2010).
Cited by
Efficient Processing of Deep Neural Networks: A Tutorial and Survey
Sze, V., Chen, Y.-H., Yang, T.-J., & Emer, J. S. (2017) · Proceedings of the IEEE · doi:10.1109/jproc.2017.2761740 · Cited by 4,215
Memristors 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
In-memory computing with resistive switching devices
Ielmini, D., & Wong, H.-S. P. (2018) · Nature Electronics · doi:10.1038/s41928-018-0092-2 · Cited by 2,298
“…can be adopted for a broad range of problems, including image compression 91 , sparse coding 93 , and implementation of artificial neural networks (ANNs), where G ij has the meaning of a synaptic weight, V j is a pre-synaptic spike amplitude, and I i is the input signal to the i th neuron 69,70 .”
BackgroundMethodMemory 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
Towards spike-based machine intelligence with neuromorphic computing
Roy, K., Jaiswal, A., & Panda, P. (2019) · Nature · doi:10.1038/s41586-019-1677-2 · Cited by 2,190
Metadata
- Authors
- Prezioso, M., Merrikh-Bayat, F., Hoskins, B. D., Adam, G. C., Likharev, K. K., & Strukov, D. B.
- Affiliations
- University of California, Santa Barbara (US)
- Stony Brook University (US)
- Corresponding author
- Prezioso, M.
- Strukov, D. B.
- Venue
- Nature
- Published
- 2015-05
- Volume
- 521
- Issue
- 7550
- Pages
- 61-64
- ISSN
- 0028-0836 (ISSN-L)
- 1476-4687
- OpenAlex
- W1542981317
- Semantic Scholar
- b138688d04eac3153bc4112babecc0106059c220
- Cited by
- 3,001
- References
- 33
- Citations by year
- 2026: 188 · 2025: 254 · 2024: 253 · 2023: 304 · 2022: 319 · 2021: 295 · 2020: 352 · 2019: 323 · 2018: 320 · 2017: 209 · 2016: 142 · 2015: 42
- FWCI
- 114.76
- Type
- journal article
- Language
- en
- Open access
- green
- arxiv.org
- Licence
- www.springer.com/tdm (tdm)
- www.springer.com/tdm (unspecified)
- Locations
- Nature (publishedVersion)
- PubMed (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- RePEc: Research Papers in Economics (submittedVersion)
- arXiv (Cornell University)
- Funders
- Defense Advanced Research Projects Agency
- Multidisciplinary University Research Initiative
- Air Force Office of Scientific Research
- 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 Applications
- Keywords
- metal-oxide memristors, memristor crossbar, device variability, single-layer perceptron, handwritten character recognition, artificial synapses, memristive computing, current-voltage characteristics
- MeSH terms
- Algorithms, Electronics, Engineering, Equipment Design, Humans, Metals, Models, Neurological, Oxides, Semiconductors, Synapses, Transistors, Electronic, Neural Networks, Computer, Biomimetics, Nanotechnology
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.