{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,17]],"date-time":"2026-09-17T17:47:55Z","timestamp":1789667275182,"version":"4.0.1"},"reference-count":37,"publisher":"Wiley","issue":"8","license":[{"start":{"date-parts":[[2022,1,17]],"date-time":"2022-01-17T00:00:00Z","timestamp":1642377600000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100008982","name":"National Science Foundation","doi-asserted-by":"publisher","award":["CCF-1900675"],"award-info":[{"award-number":["CCF-1900675"]}],"id":[{"id":"10.13039\/501100008982","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["advanced.onlinelibrary.wiley.com"],"crossmark-restriction":true},"short-container-title":["Advanced Intelligent Systems"],"published-print":{"date-parts":[[2022,8]]},"abstract":"<jats:sec><jats:label\/><jats:p>In analog in\u2010memory computing systems based on nonvolatile memories such as resistive random\u2010access memory (RRAM), neural network models are often trained offline and then the weights are programmed onto memory devices as conductance values. The programmed weight values inevitably deviate from the target values during the programming process. This effect can be pronounced for emerging memories such as RRAM, PcRAM, and MRAM due to the stochastic nature during programming. Unlike noise, these weight deviations do not change during inference. The performance of neural network models is investigated against this programming variation under realistic system limitations, including limited device on\/off ratios, memory array size, analog\u2010to\u2010digital converter (ADC) characteristics, and signed weight representations. Approaches to mitigate such device and circuit nonidealities through architecture\u2010aware training are also evaluated. The effectiveness of variation injection during training to improve the inference robustness, as well as the effects of different neural network training parameters such as learning rate schedule, will be discussed.<\/jats:p><\/jats:sec>","DOI":"10.1002\/aisy.202100199","type":"journal-article","created":{"date-parts":[[2022,1,17]],"date-time":"2022-01-17T04:10:37Z","timestamp":1642392637000},"update-policy":"https:\/\/doi.org\/10.1002\/crossmark_policy","source":"Crossref","is-referenced-by-count":29,"title":["Device Variation Effects on Neural Network Inference Accuracy in Analog In\u2010Memory Computing Systems"],"prefix":"10.1002","volume":"4","author":[{"given":"Qiwen","family":"Wang","sequence":"first","affiliation":[{"name":"Department of Electrical Engineering and Computer Science University of Michigan  Ann Arbor MI 48109 USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yongmo","family":"Park","sequence":"additional","affiliation":[{"name":"Department of Electrical Engineering and Computer Science University of Michigan  Ann Arbor MI 48109 USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-4731-1976","authenticated-orcid":false,"given":"Wei D.","family":"Lu","sequence":"additional","affiliation":[{"name":"Department of Electrical Engineering and Computer Science University of Michigan  Ann Arbor MI 48109 USA"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"311","published-online":{"date-parts":[[2022,1,17]]},"reference":[{"key":"e_1_2_11_2_1","unstructured":"E.Strubell A.Ganesh A.McCallum inAAAI 2020\u201434th AAAI Conf. Artif. Intell. New York2020 pp.1393\u201313696."},{"key":"e_1_2_11_3_1","doi-asserted-by":"publisher","DOI":"10.1109\/MCAS.2021.3092533"},{"key":"e_1_2_11_4_1","doi-asserted-by":"publisher","DOI":"10.1016\/B978-0-08-102782-0.00009-5"},{"key":"e_1_2_11_5_1","doi-asserted-by":"crossref","unstructured":"X.Peng S.Huang Y.Luo X.Sun S.Yu Technical Digest \u2010 Int. Electron Devices meeting IEDM 2019 San Francisco2019 pp.32.5.1\u201332.5.4.","DOI":"10.1109\/IEDM19573.2019.8993491"},{"key":"e_1_2_11_6_1","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-020-16108-9"},{"key":"e_1_2_11_7_1","doi-asserted-by":"publisher","DOI":"10.1038\/s41928-019-0270-x"},{"key":"e_1_2_11_8_1","doi-asserted-by":"publisher","DOI":"10.1038\/s41586-020-1942-4"},{"key":"e_1_2_11_9_1","doi-asserted-by":"crossref","unstructured":"O.Krestinskaya B.Choubey A. P.James Sci. Rep.2020 10 5838.","DOI":"10.1038\/s41598-020-62676-7"},{"key":"e_1_2_11_10_1","doi-asserted-by":"crossref","unstructured":"O.Krestinskaya A.Irmanova A. P.James in2019 IEEE Int. Symp. Circuits Syst. Sapporo 2019 pp. 1\u20135.","DOI":"10.1109\/ISCAS.2019.8702245"},{"key":"e_1_2_11_11_1","unstructured":"B.Murmann Boris Murmann: ADC Performance Survey https:\/\/web.stanford.edu\/~murmann\/adcsurvey.html 2021."},{"key":"e_1_2_11_12_1","doi-asserted-by":"publisher","DOI":"10.1109\/TNANO.2017.2784364"},{"key":"e_1_2_11_13_1","doi-asserted-by":"crossref","unstructured":"Q.Wang X.Wang S. H.Lee F.\u2010H.Meng W. D.Lu inTechnical Digest \u2010 Int. Electron Devices Meeting IEDM 2019 San Francisco2019 pp.14.4.1\u201314.4.4.","DOI":"10.1109\/IEDM19573.2019.8993641"},{"key":"e_1_2_11_14_1","doi-asserted-by":"crossref","unstructured":"W.Ma F.Cai C.Du Y.Jeong M.Zidan W. D.Lu inTechnical Digest\u2014Int. Electron Devices Meeting IEDM 2017 San Francisco2017 p.16.7.1.","DOI":"10.1109\/IEDM.2016.7838434"},{"key":"e_1_2_11_15_1","unstructured":"Q.Wang Y.Park W. D.Lu in2021 IEEE Int. Symp. Circuits Systems IEEE Piscataway NJ2021 pp.1\u20135."},{"key":"e_1_2_11_16_1","doi-asserted-by":"crossref","unstructured":"X.Wang Q.Wang F. H.Meng S. H.Lee W. D.Lu inProc.\u20142020 IEEE Int. Conf. Artificial Intelligence Circuits and Systems AICAS 2020 Genoa2020 pp.141\u2013144.","DOI":"10.1109\/AICAS48895.2020.9073942"},{"key":"e_1_2_11_17_1","unstructured":"R.Krishnamoorthi (Preprint) arXiv:1806.08342 v1 unpublished2018."},{"key":"e_1_2_11_18_1","unstructured":"K.Simonyan A.Zisserman in3rd Int. Conf. on Learning Representations ICLR 2015 San Diego2015."},{"key":"e_1_2_11_19_1","unstructured":"B. E.Jonsson inIMEKO TC4 Int. Workshop ADC Model. Test. Data Convert. Anal. Des. 2011 IWADC 2011 Orvieto2011 pp.132\u2013137."},{"key":"e_1_2_11_20_1","doi-asserted-by":"crossref","unstructured":"S.Zagoruyko N.Komodakis inProc. Br. Mach. Vis. Conf. 2016 York2016 pp.87.1\u201387.12.","DOI":"10.5244\/C.30.87"},{"key":"e_1_2_11_21_1","unstructured":"L.\u2010H.Tsai S.\u2010C.Chang Y.\u2010T.Chen J.\u2010Y.Pan W.Wei D.\u2010C.Juan (Preprint) arXiv:2007.03230 v2 unpublished2020."},{"key":"e_1_2_11_22_1","doi-asserted-by":"publisher","DOI":"10.1038\/s41928-018-0100-6"},{"key":"e_1_2_11_23_1","doi-asserted-by":"crossref","unstructured":"X.Guo F. M.Bayat M.Prezioso Y.Chen B.Nguyen N.Do D. B.Strukov in2017 IEEE Custom Integrated Circuits Conf. Austin2017 pp.1\u20134.","DOI":"10.1109\/CICC.2017.7993628"},{"key":"e_1_2_11_24_1","doi-asserted-by":"crossref","unstructured":"Y.Cai E. F.Haratsch O.Mutlu K.Mai inDesign Automation & Test in Europe Conf. and Exhibition (DATE) 2013 New Jersey2013 pp.1285\u20131290.","DOI":"10.7873\/DATE.2013.266"},{"key":"e_1_2_11_25_1","doi-asserted-by":"publisher","DOI":"10.1109\/72.238328"},{"key":"e_1_2_11_26_1","unstructured":"R. D.Clay C. H.Sequin inJCNN Int. Jt. Conf. Neural Networks Baltimore1992 pp.769\u2013774."},{"key":"e_1_2_11_27_1","doi-asserted-by":"publisher","DOI":"10.1162\/neco.1996.8.3.643"},{"key":"e_1_2_11_28_1","unstructured":"D. P.Kingma T.Salimans M.Welling inProc. 28th Int. Conf. Neural Inf. Process. Syst. NIPS 2015 Montr\u00e9al2015 pp.2575\u20132583."},{"key":"e_1_2_11_29_1","first-page":"5110","volume-title":"Advances in Neural Information Processing Systems","author":"Noh H.","year":"2017"},{"key":"e_1_2_11_30_1","doi-asserted-by":"publisher","DOI":"10.1109\/TVLSI.2019.2893256"},{"key":"e_1_2_11_31_1","doi-asserted-by":"publisher","DOI":"10.1109\/JSSC.2017.2712626"},{"key":"e_1_2_11_32_1","doi-asserted-by":"crossref","unstructured":"M.Klachko M. R.Mahmoodi D.Strukov inProc. Int. Joint Conf. Neural Networks IJCNN'19 Budapest2019 pp.1\u20138.","DOI":"10.1109\/IJCNN.2019.8851966"},{"key":"e_1_2_11_33_1","doi-asserted-by":"crossref","unstructured":"A. S.Rekhi B.Zimmer N.Nedovic N.Liu R.Venkatesan M.Wang B.Khailany W. J.Dally C. T.Gray inProc. 56th Annual Design Automation Conf. DAC 2019 Las Vegas 2019 pp.1\u20136.","DOI":"10.1145\/3316781.3317770"},{"key":"e_1_2_11_34_1","doi-asserted-by":"crossref","unstructured":"Y.Boo S.Shin W.Sung in2020 IEEE Work. Signal Process. Syst. Coimbra2020 pp.1\u20136.","DOI":"10.1109\/SiPS50750.2020.9195245"},{"key":"e_1_2_11_35_1","doi-asserted-by":"crossref","unstructured":"L. N.Smith inProc.\u20142017 IEEE Winter Conf. Appl. Comput. Vision WACV 2017 Santa Rosa2017 pp.464\u2013472.","DOI":"10.1109\/WACV.2017.58"},{"key":"e_1_2_11_36_1","doi-asserted-by":"publisher","DOI":"10.1109\/TCSI.2021.3079980"},{"key":"e_1_2_11_37_1","doi-asserted-by":"crossref","unstructured":"S.\u2010S.Sheu M.\u2010F.Chang K.\u2010F.Lin C.\u2010W.Wu Y.\u2010S.Chen P.\u2010F.Chiu C.\u2010C.Kuo Y.\u2010S.Yang P.\u2010C.Chiang W.\u2010P.Lin C.\u2010H.Lin H.\u2010Y.Lee P.\u2010Y.Gu S.\u2010M.Wang F. T.Chen K.\u2010L.Su C.\u2010H.Lien K.\u2010H.Cheng H.\u2010T.Wu T.\u2010K.Ku M.\u2010J.Kao M.\u2010J.Tsai in2011 IEEE Int. Solid-State Circuits Conf. ISSCC 2011 San Francisco2011 pp.200\u2013202.","DOI":"10.1109\/ISSCC.2011.5746281"},{"key":"e_1_2_11_38_1","unstructured":"N.Papandreou H.Pozidis A.Pantazi A.Sebastian M.Breitwisch C.Lam E.Eleftheriou in2011 IEEE Int. Symp. Circuits Systems Rio de Janeiro2011 pp.329\u2013332."}],"container-title":["Advanced Intelligent Systems"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/onlinelibrary.wiley.com\/doi\/pdf\/10.1002\/aisy.202100199","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/onlinelibrary.wiley.com\/doi\/full-xml\/10.1002\/aisy.202100199","content-type":"application\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/advanced.onlinelibrary.wiley.com\/doi\/pdf\/10.1002\/aisy.202100199","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,7]],"date-time":"2025-10-07T15:43:32Z","timestamp":1759851812000},"score":1,"resource":{"primary":{"URL":"https:\/\/advanced.onlinelibrary.wiley.com\/doi\/10.1002\/aisy.202100199"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2022,1,17]]},"references-count":37,"journal-issue":{"issue":"8","published-print":{"date-parts":[[2022,8]]}},"alternative-id":["10.1002\/aisy.202100199"],"URL":"https:\/\/doi.org\/10.1002\/aisy.202100199","archive":["Portico"],"relation":{},"ISSN":["2640-4567","2640-4567"],"issn-type":[{"value":"2640-4567","type":"print"},{"value":"2640-4567","type":"electronic"}],"subject":[],"published":{"date-parts":[[2022,1,17]]},"assertion":[{"value":"2021-10-04","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2022-01-17","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}],"article-number":"2100199"}}