A Probabilistic Bayesian Machine Based on a Diffusive Memristor Crossbar Array for Long Sequence Inference

Yining Jiang , Hanzhi Ma , Yongqin Bai , Xun Han , Ye Shi , Jose Schutt-Aine , Yang Xu , Er-Ping Li

Engineering ›› : 202607011

PDF (3621KB)
Engineering ›› :202607011 DOI: 10.1016/j.eng.2026.07.011
research-article
A Probabilistic Bayesian Machine Based on a Diffusive Memristor Crossbar Array for Long Sequence Inference
Author information +
History +
PDF (3621KB)

Abstract

The current Von-Neumann architecture cannot support emerging applications in artificial intelligence, necessitating a new computing paradigm that can solve complex tasks. Memristor-based near-memory computing demonstrates the potential beyond Von-Neumann computers. However, due to device limita- tions, memristor-based hardware with bulky peripheral circuits cannot handle complex tasks and is sen- sitive to noise. In this study, we leverage the physical stochasticity of a diffusive device and the inherent parallelism of the crossbar structure to implement a true stochastic Bayesian machine based on an Au/Ag/ Al2O3/Pt/Ti memristor crossbar array circuit. The proposed system supports long probabilistic sequence inference with superior soft-error robustness and enables probabilistic hardware to handle high-feature tasks such as image recognition for the first time. We have constructed a diffusive memristor device with the corresponding compact model and measured the signal-transmission characteristics of the crossbar array prototype circuit, revealing severe signal distortion due to parasitic effects. Moreover, we propose a signal-reconstruction method that effectively improves the performance of Bayesian machine circuits. Our work points to new directions for exploring circuit design and large-scale integration methods in probabilistic computing hardware.

Keywords

Probabilistic computing / Diffusive memristor / Bayesian machine / Long sequence inference / Soft-error robustness

Cite this article

Download citation ▾
Yining Jiang, Hanzhi Ma, Yongqin Bai, Xun Han, Ye Shi, Jose Schutt-Aine, Yang Xu, Er-Ping Li. A Probabilistic Bayesian Machine Based on a Diffusive Memristor Crossbar Array for Long Sequence Inference. Engineering 202607011 DOI:10.1016/j.eng.2026.07.011

登录浏览全文

4963

注册一个新账户 忘记密码

References

Publishing order | Descend order by publishing year | Descend order by cited within
[1]

Zidan MA, Strachan JP, Lu WD . The future of electronics based on memristive systems. Nat Electron 2018;1(1):22-9.
[2]

Gao Y, Ji S, Han X, Dai Q . Hypergraph computation. Engineering 2024; 40:188-201.
[3]

Schuman CD, Kulkarni SR, Parsa M, Mitchell JP, Date P, Kay B . Opportunities for neuromorphic computing algorithms and applications. Nat Comput Sci 2022; 2(1):10-9.
[4]

Zheng Q, Liu H, Zhang X, Yan C, Cao X, Gong T, et al. Machine memory intelligence: inspired by human memory mechanisms. Engineering 2025; 55:24-35.
[5]

Xu D, Ma Y, Jin G, Cao L . Intelligent photonics: a disruptive technology to shape the present and redefine the future. Engineering 2025; 46:186-213.
[6]

Kim H, Sohn W, Lee JH, Yang SJ, Shim YS, Moon CW, et al. Recent advances in MXene materials for resistive switching memory devices. J Korean Ceram Soc 2025; 63:171-87.
[7]

Kim H, Sohn W, Hilal M, Cai Z, Hou J, Shin W, et al. Exploring the potential of halide perovskites in designing resistive switching memory systems for next—generation electronic devices. J Korean Ceram Soc 2025; 62(6):1099-120.
[8]

Yue W, Zhang T, Jing Z, Wu K, Yang Y, Yang Z, et al. A scalable universal Ising machine based on interaction—centric storage and compute—in—memory. Nat Electron 2024; 7(10):904—13.
[9]

Zhang X, Jiang C, Yin J, Zhu D, Wang S, Li S, et al. Experimental realization of physical unclonable function chip utilizing spintronic memories. Engineering 2025; 49:141-8.
[10]

Lin J, Zhang S, Li S, Xu Y, Li X, Duan W, et al. Initialization—free programmable spin—logic gate in a single spin—orbit torque device. Engineering 2025; 51:215—20.
[11]

Tao T, Ma H, Chen Q, Gu ZM, Jin H, Ahmed M, et al. Circuit modeling for RRAM—based neuromorphic chip crossbar array with and without write—verify scheme. IEEE Trans Circuits Syst I Regul Pap 2021; 68(5):1906—16.
[12]

Lin R, Yao J, Wang Z, Chan CT, Tsai DP . Nonlinear meta—devices: from plasmonic to dielectric. Engineering 2025; 45:15-24.
[13]

Kim K, Song MS, Hwang H, Hwang S, Kim H . A comprehensive review of advanced trends: from artificial synapses to neuromorphic systems with consideration of non—ideal effects. Front Neurosci 2024; 18:1279708.
[14]

Duan X, Cao Z, Gao K, Yan W, Sun S, Zhou G, et al. Memristor—based neuromorphic chips. Adv Mater 2024; 36(14):2310704.
[15]

Kumar S, Yadav D, Ramesh R, Stathopoulos S, Tsiamis A, Prodromakis T . Electrical analysis of atomic layer deposited thin HfO2 and HfO2/Ta2O5—based memristive devices. IEEE Trans Electron Dev 2025; 72(4):1780—7.
[16]

Yang C, Wang H, Wang K, Cao Z, Ren F, Zhou G, et al. Silk fibroin—based biomemristors for bionic artificial intelligence robot applications. ACS Nano 2025; 19(18):17173—98.
[17]

Shin W, Jang S, Kim H . Recent advances in resistive switching memory devices based on diverse material platforms for next generation electronics. J Sci Adv Mater Devices 2025; 10(3):100960.
[18]

Kim H. 2D halide perovskites in modern electronics and their impact on high—performance resistive—switching memory device and field—effect transistors. J Korean Ceram Soc 2025; 63:1-12.
[19]

Woo KS, Kim J, Han J, Kim W, Jang YH, Hwang CS . Probabilistic computing using Cu0.1Te0.9/HfO2/Pt diffusive memristors. Nat Commun 2022; 13(1):5762.
[20]

Camsari KY, Faria R, Sutton BM, Datta S . Stochastic p—bits for invertible logic. Phys Rev X 2017; 7(3):031014.
[21]

Liu S, Xiao TP, Kwon J, Debusschere BJ, Agarwal S, Incorvia JAC, et al. Bayesian neural networks using magnetic tunnel junction—based probabilistic in—memory computing. Front Nanotechnol 2022; 4:1021943.
[22]

Camsari KY, Chowdhury S, Datta S . Scalable emulation of sign—problem—free Hamiltonians with room—temperature p—bits. Phys Rev Appl 2019; 12(3):034061.
[23]

Neal RM . Connectionist learning of belief networks. Artif Intell 1992; 56(1):71-113.
[24]

Liu S, Mingas G, Bouganis CS . An unbiased MCMC FPGA—based accelerator in the land of custom precision arithmetic. IEEE Trans Comput 2017; 66(5):745-58.
[25]

Ko GG, Chai Y, Donato M, Whatmough PN, Tambe T, Rutenbar RA, et al. A 3 mm2 programmable Bayesian inference accelerator for unsupervised machine perception using parallel Gibbs sampling in 16 nm. In: Proceedings of the 2020 IEEE Symposium on VLSI Circuits; 2020 Jun 16—19; Honolulu, HI, USA. Piscataway: Institute of Electrical and Electronics Engineers; 2020. p. 1-2.
[26]

Ghahramani Z. Probabilistic machine learning and artificial intelligence. Nature 2015; 521(7553):452-9.
[27]

Baek IK, Lee SH, Jang YH, Park H, Kim J, Cheong S, et al. Implementation of Bayesian networks and Bayesian inference using a Cu0.1Te0.9/HfO2/Pt threshold switching memristor. Nanoscale Adv 2024; 6(11):2892-902.
[28]

Harabi KE, Hirtzlin T, Turck C, Vianello E, Laurent R, Droulez J, et al. A memristor—based Bayesian machine. Nat Electron 2022; 6:52-63.
[29]

Pan W, Zheng J, Wang L, Luo Y . A future perspective on in—sensor computing. Engineering 2022; 14:19-21.
[30]

Thomas SA, Vohra SK, Kumar R, Sharma R, Das DM . Analysis of parasitic effects in a crossbar in CMOS based neuromorphic system for pattern recognition using memristive synapses. IEEE Trans Nanotechnol 2022; 21:380—9.
[31]

Yi S, Kendall JD, Williams RS, Kumar S . Activity—difference training of deep neural networks using memristor crossbars. Nat Electron 2023; 6:45-51.
[32]

Li EP, Ma H, Ahmed M, Tao T, Gu Z, Chen M, et al. An electromagnetic perspective of artificial intelligence neuromorphic chips. Electromagn Sci 2023;1(3):1-18.
[33]

Tao T, Li D, Ma H, Li Y, Tan S, Liu E, et al. A new pre—conditioned STDP rule and its hardware implementation in neuromorphic crossbar array. Neurocomputing 2023; 557:126682.
[34]

Kumar S, Kumbhar DD, Park JH, Kamat RK, Dongale TD, Mukherjee S . Y2O3—based crossbar array for analog and neuromorphic computation. IEEE Trans Electron Dev 2023; 70(2):473—7.
[35]

Vodenicarevic D, Locatelli N, Mizrahi A, Friedman JS, Vincent AF, Romera M, et al. Low—energy truly random number generation with superparamagnetic tunnel junctions for unconventional computing. Phys Rev Appl 2017; 8(5):054045.
[36]

Groner MD, Fabreguette FH, Elam JW, George SM . Low—temperature Al2O3 atomic layer deposition. Chem Mater 2004; 16(4):639—45.
[37]

Zhu J, Zhang T, Yang Y, Huang R . A comprehensive review on emerging artificial neuromorphic devices. Appl Phys Rev 2020; 7(1):011312.
[38]

Wang Z, Joshi S, Savel’ev SE, Jiang H, Midya R, Lin P, et al. Memristors with diffusive dynamics as synaptic emulators for neuromorphic computing. Nat Mater 2017; 16(1):101-8.
[39]

Zhang X, Wang W, Liu Q, Zhao X, Wei J, Cao R, et al. An artificial neuron based on a threshold switching memristor. IEEE Electron Device Lett 2018; 39(2):308—11.
[40]

Gao B, Lin B, Pang Y, Xu F, Lu Y, Chiu YC, et al. Concealable physically unclonable function chip with a memristor array. Sci Adv 2022; 8(24):eabn7753.
[41]

Yoon JH, Wang Z, Kim KM, Wu H, Ravichandran V, Xia Q, et al. An artificial nociceptor based on a diffusive memristor. Nat Commun 2018; 9(1):417.
[42]

Zhang R, Jiang H, Wang ZR, Lin P, Zhuo Y, Holcomb D, et al. Nanoscale diffusive memristor crossbars as physical unclonable functions. Nanoscale 2018; 10(6):2721—6.
[43]

Jiang Y, Ma H, Bai Y, Han X, Xu Y, Li EP . Circuit modeling and analysis of memristor crossbar array—based SNN with spike—mapping STDP implementation. IEEE Trans Very Large Scale Integr Syst 2026; 34:193-204.
[44]

Chen A, Hu J, Ma H, Jiang Y, Yu B . Differentiable distribution model of stochastic volatile memristor—based neuron. IEEE Trans Electron Dev 2025; 72(4):1709—14.
[45]

Woo KS, Kim J, Han J, Choi JM, Kim W, Hwang CS . A high—speed true random number generator based on a CuxTe1−x diffusive memristor. Adv Intell Syst 2021; 3(7):2100062.
[46]

Woo KS, Wang Y, Kim J, Kim Y, Kwon YJ, Yoon JH, et al. A true random number generator using threshold—witching—based memristors in an efficient circuit design. Adv Electron Mater 2019; 5(2):1800543.
[47]

Woo KS, Wang Y, Kim Y, Kim J, Kim W, Hwang CS . Hardware—based security: a combination of a volatile—memristor—based true random—number generator and a nonlinear—feedback shift register for high—speed encryption. Adv Electron Mater 2020; 6(5):2070022.
[48]

Li Z, Wang J, Xu L, Wang L, Shang H, Ying H, et al. Achieving reliable and ultrafast memristors via artificial filaments in silk fibroin. Adv Mater 2024; 36(4):2308843.
[49]

Wang W, Laudato M, Ambrosi E, Bricalli A, Covi E, Lin YH, et al. Volatile resistive switching memory based on ag ion drift/diffusion part I: numerical modeling. IEEE Trans Electron Dev 2019; 66(9):3795-801.
[50]

Wang W, Laudato M, Ambrosi E, Bricalli A, Covi E, Lin YH, et al. Volatile resistive switching memory based on ag ion drift/diffusion—part II: compact modeling. IEEE Trans Electron Dev 2019; 66(9):3802—8.
[51]

Chen PY, Yu S . Compact modeling of RRAM devices and its applications in 1T1R and 1S1R array design. IEEE Trans Electron Dev 2015; 62(12):4022—8.
[52]

García—Redondo F, Gowers RP, Crespo—Yepes A, López—Vallejo M, Jiang L . SPICE compact modeling of bipolar/unipolar memristor switching governed by electrical thresholds. IEEE Trans Circuits Syst I Regul Pap 2016; 63(8):1255-64.
[53]

Liao Y, Gao B, Xu F, Yao P, Chen J, Zhang W, et al. A compact model of analog RRAM with device and array nonideal effects for neuromorphic systems. IEEE Trans Electron Dev 2020; 67(4):1593—9.
[54]

Wang M, Wang W, Leow WR, Wan C, Chen G, Zeng Y, et al. Enhancing the matrix addressing of flexible sensory arrays by a highly nonlinear threshold switch. Adv Mater 2018; 30(33):1802516.
[55]

Chen Y, Kang Y, Hao H, Xie X, Zeng J, Xu T, et al. All two—dimensional integration—type optoelectronic synapse mimicking visual attention mechanism for multi—target recognition. Adv Funct Mater 2023; 33(6):2209781.
[56]

Zhao M, Gao B, Tang J, Qian H, Wu H . Reliability of analog resistive switching memory for neuromorphic computing. Appl Phys Rev 2020; 7(1):011301.
[57]

Maor G, Zeng X, Wang Z, Hu Y . An FPGA implementation of stochastic computing—based LSTM. In: Proceedings of the 2019 IEEE 37th International Conference on Computer Design; 2019 Nov 17—20; Abu Dhabi, United Arab Emirates. Piscataway: Institute of Electrical and Electronics Engineers; 2019. p. 38-46.
[58]

Chen A. A comprehensive crossbar array model with solutions for line resistance and nonlinear device characteristics. IEEE Trans Electron Dev 2013; 60(4):1318—26.
[59]

Agarwal K, Sylvester D, Blaauw D . Modeling and analysis of crosstalk noise in coupled RLC interconnects. IEEE Trans Comput Aided Des Integrated Circ Syst 2006; 25(5):892-901.
[60]

Davis JA, Meindl JD . Compact distributed RLC interconnect models. I. Single line transient, time delay, and overshoot expressions. IEEE Trans Electron Dev 2000; 47(11):2068-77.
[61]

Davis JA, Meindl JD . Compact distributed RLC interconnect models—part II: coupled line transient expressions and peak crosstalk in multilevel networks. IEEE Trans Electron Dev 2000; 47(11):2078-87.
[62]

Fouda ME, Eltawil AM, Kurdahi F . Modeling and analysis of passive switching crossbar arrays. IEEE Trans Circuits Syst I Regul Pap 2018; 65(1):270-82.
[63]

Lee W, Kim J . Accuracy investigation of a neuromorphic machine learning system due to electromagnetic noises using PEEC model. IEEE Trans Compon Packaging Manuf Technol 2019; 9(10):2066-78.
[64]

Starzyk JA, Basawaraj. Memristor crossbar architecture for synchronous neural networks. IEEE Trans Circuits Syst 2014; I(61):2390-401.
[65]

Jiao B, Qiao J, Jia S, Liu R, Wei X, Yun S, et al. Low stress TSV arrays for high—density interconnection. Engineering 2024; 38:201-8.
[66]

Shomalnasab G, Zhang L . New analytic model of coupling and substrate capacitance in nanometer technologies. IEEE Trans Very Large Scale Integr Syst 2015; 23(7):1268-80.
[67]

Dragomiretskiy K, Zosso D . Variational mode decomposition. IEEE Trans Signal Process 2014; 62(3):531—44.
[68]

Saini M, Payal SU . An effective and robust framework for ocular artifact removal from single—channel EEG signal based on variational mode decomposition. IEEE Sens J 2020; 20(1):369-76.
[69]

Shahbakhti M, Beiramvand M, Nazari M, Broniec—Wojcik A, Augustyniak P, Rodrigues AS, et al. VME—DWT: an efficient algorithm for detection and elimination of eye blink from short segments of single EEG channel. IEEE Trans Neural Syst Rehabil Eng 2021; 29:408—17.
[70]

Jiang Y, Wu D, Cao J, Jiang L, Zhang S, Wang D . Eyeblink detection algorithm based on joint optimization of VME and morphological feature extraction. IEEE Sens J 2023; 23(18):21374—84.
[71]

Nazari M, Sakhaei SM . Variational mode extraction: a new efficient method to derive respiratory signals from ECG. IEEE J Biomed Health Inform 2018; 22(4):1059-67.
[72]

Sahani M, Dash PK . FPGA—based deep convolutional neural network of process adaptive VMD data with online sequential RVFLN for power quality events recognition. IEEE Trans Power Electron 2021; 36(4):4006—15.
[73]

Sahani M, Dash PK . FPGA—based semisupervised multifusion RDCNN of process robust VMD data with online kernel RVFLN for power quality events recognition. IEEE Trans Neural Netw Learn Syst 2022; 33(2):515-27.
[74]

Wang Z, Xu Z, He J, Delingette H, Fan J . Long short—term memory neural equalizer. IEEE Trans Signal Power Integr 2023; 2:13-22.
[75]

Kim H, Im IH, Hyun D, Hilal M, Cai Z, Yang SJ, et al. Stretchable resistive switching memory devices for wearable systems. J Korean Ceram Soc 2025; 62(3):397-411.
[76]

Roy K, Jaiswal A, Panda P . Towards spike—based machine intelligence with neuromorphic computing. Nature 2019; 575(7784):607-17.
[77]

Rathi N, Chakraborty I, Kosta A, Sengupta A, Ankit A, Panda P, et al. Exploring neuromorphic computing based on spiking neural networks: algorithms to hardware. ACM Comput Surv 2023; 55(12):1-49.
[78]

Jiang J, Yang L, Yang S, Zhang L . A deep learning—based framework for environment—adaptive navigation of size—adaptable microswarms. Engineering 2025; 53:130—8.
[79]

Kim H. Halide perovskites for next generation electronics integrating resistive switching memory and gas sensing. J Korean Ceram Soc 2025; 62(5):845—60.
PDF (3621KB)

81

Accesses

0

Citation

Detail
Sections
Recommended

/

〈 〉