{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,2,20]],"date-time":"2026-02-20T07:17:07Z","timestamp":1771571827800,"version":"3.50.1"},"reference-count":19,"publisher":"IOP Publishing","issue":"3","license":[{"start":{"date-parts":[[2023,8,29]],"date-time":"2023-08-29T00:00:00Z","timestamp":1693267200000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2023,8,29]],"date-time":"2023-08-29T00:00:00Z","timestamp":1693267200000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/iopscience.iop.org\/info\/page\/text-and-data-mining"}],"funder":[{"DOI":"10.13039\/100000015","name":"U.S. Department of Energy","doi-asserted-by":"crossref","award":["#DE-SC0019273"],"award-info":[{"award-number":["#DE-SC0019273"]}],"id":[{"id":"10.13039\/100000015","id-type":"DOI","asserted-by":"crossref"}]},{"DOI":"10.13039\/100000002","name":"National Institutes of Health","doi-asserted-by":"crossref","award":["DP2 EB030992"],"award-info":[{"award-number":["DP2 EB030992"]}],"id":[{"id":"10.13039\/100000002","id-type":"DOI","asserted-by":"crossref"}]},{"name":"National Science Foundation","award":["ECCS-1542148"],"award-info":[{"award-number":["ECCS-1542148"]}]},{"DOI":"10.13039\/100007297","name":"Office of Naval Research Global","doi-asserted-by":"crossref","award":["N000142012405"],"award-info":[{"award-number":["N000142012405"]}],"id":[{"id":"10.13039\/100007297","id-type":"DOI","asserted-by":"crossref"}]}],"content-domain":{"domain":["iopscience.iop.org"],"crossmark-restriction":false},"short-container-title":["Neuromorph. Comput. Eng."],"published-print":{"date-parts":[[2023,9,1]]},"abstract":"<jats:title>Abstract<\/jats:title>\n               <jats:p>In-memory computing with emerging non-volatile memory devices (eNVMs) has shown promising results in accelerating matrix-vector multiplications. However, activation function calculations are still being implemented with general processors or large and complex neuron peripheral circuits. Here, we present the integration of Ag-based conductive bridge random access memory (Ag-CBRAM) crossbar arrays with Mott rectified linear unit (ReLU) activation neurons for scalable, energy and area-efficient hardware (HW) implementation of deep neural networks. We develop Ag-CBRAM devices that can achieve a high ON\/OFF ratio and multi-level programmability. Compact and energy-efficient Mott ReLU neuron devices implementing ReLU activation function are directly connected to the columns of Ag-CBRAM crossbars to compute the output from the weighted sum current. We implement convolution filters and activations for VGG-16 using our integrated HW and demonstrate the successful generation of feature maps for CIFAR-10 images in HW. Our approach paves a new way toward building a highly compact and energy-efficient eNVMs-based in-memory computing system.<\/jats:p>","DOI":"10.1088\/2634-4386\/aceea9","type":"journal-article","created":{"date-parts":[[2023,8,9]],"date-time":"2023-08-09T22:32:06Z","timestamp":1691620326000},"page":"034007","update-policy":"https:\/\/doi.org\/10.1088\/crossmark-policy","source":"Crossref","is-referenced-by-count":5,"title":["Integration of Ag-CBRAM crossbars and Mott ReLU neurons for efficient implementation of deep neural networks in hardware"],"prefix":"10.1088","volume":"3","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-3246-6994","authenticated-orcid":true,"given":"Yuhan","family":"Shi","sequence":"first","affiliation":[]},{"given":"Sangheon","family":"Oh","sequence":"additional","affiliation":[]},{"ORCID":"https:\/\/orcid.org\/0000-0002-7680-9709","authenticated-orcid":true,"given":"Jaeseoung","family":"Park","sequence":"additional","affiliation":[]},{"given":"Javier del","family":"Valle","sequence":"additional","affiliation":[]},{"given":"Pavel","family":"Salev","sequence":"additional","affiliation":[]},{"given":"Ivan K","family":"Schuller","sequence":"additional","affiliation":[]},{"given":"Duygu","family":"Kuzum","sequence":"additional","affiliation":[]}],"member":"266","published-online":{"date-parts":[[2023,8,29]]},"reference":[{"key":"nceaceea9bib1","doi-asserted-by":"publisher","first-page":"436","DOI":"10.1038\/nature14539","article-title":"Deep learning","volume":"521","author":"LeCun","year":"2015","journal-title":"Nature"},{"key":"nceaceea9bib2","first-page":"1","article-title":"Vanishing gradient mitigation with deep learning neural network optimization","author":"Tan","year":"2019"},{"key":"nceaceea9bib3","doi-asserted-by":"publisher","first-page":"333","DOI":"10.1038\/s41928-018-0092-2","article-title":"In-memory computing with resistive switching devices","volume":"1","author":"Ielmini","year":"2018","journal-title":"Nat. Electron."},{"key":"nceaceea9bib4","first-page":"325","article-title":"Emerging non-volatile memories: opportunities and challenges","author":"Xue","year":"2011"},{"key":"nceaceea9bib5","first-page":"22.1. 1","article-title":"Design considerations for efficient deep neural networks on processing-in-memory accelerators","author":"Yang","year":"2019"},{"key":"nceaceea9bib6","doi-asserted-by":"publisher","first-page":"719","DOI":"10.1109\/TCSI.2018.2866510","article-title":"Learning in memristive neural network architectures using analog backpropagation circuits","volume":"66","author":"Krestinskaya","year":"2018","journal-title":"IEEE Trans. Circuits Syst. I"},{"key":"nceaceea9bib7","doi-asserted-by":"publisher","first-page":"367","DOI":"10.1109\/JETCAS.2019.2911537","article-title":"Analog-to-digital conversion with reconfigurable function mapping for neural networks activation function acceleration","volume":"9","author":"Giordano","year":"2019","journal-title":"IEEE J. Emerg. Sel. Top. Circuits Syst."},{"key":"nceaceea9bib8","doi-asserted-by":"publisher","first-page":"680","DOI":"10.1038\/s41565-021-00874-8","article-title":"Energy-efficient Mott activation neuron for full-hardware implementation of neural networks","volume":"16","author":"Oh","year":"2021","journal-title":"Nat. Nanotechnol."},{"key":"nceaceea9bib9","doi-asserted-by":"publisher","first-page":"232","DOI":"10.1016\/j.neunet.2018.11.005","article-title":"A comparison of deep networks with ReLU activation function and linear spline-type methods","volume":"110","author":"Eckle","year":"2019","journal-title":"Neural Netw."},{"key":"nceaceea9bib10","doi-asserted-by":"publisher","first-page":"54","DOI":"10.1109\/MM.2019.2943047","article-title":"Monolithically integrated RRAM-and CMOS-based in-memory computing optimizations for efficient deep learning","volume":"39","author":"Yin","year":"2019","journal-title":"IEEE Micro"},{"key":"nceaceea9bib11","doi-asserted-by":"publisher","DOI":"10.1088\/1361-6463\/abc5e7","article-title":"Oxide-based filamentary RRAM for deep learning","volume":"54","author":"Zhang","year":"2020","journal-title":"J. Appl. Phys."},{"key":"nceaceea9bib12","doi-asserted-by":"publisher","first-page":"4964","DOI":"10.1039\/C4NR06922F","article-title":"Conduction mechanism of a TaOx-based selector and its application in crossbar memory arrays","volume":"7","author":"Wang","year":"2015","journal-title":"Nanoscale"},{"key":"nceaceea9bib13","first-page":"10.2. 1","article-title":"Demonstration of 3D vertical RRAM with ultra low-leakage, high-selectivity and self-compliance memory cells","author":"Luo","year":"2015"},{"key":"nceaceea9bib14","doi-asserted-by":"publisher","first-page":"4022","DOI":"10.1109\/TED.2015.2492421","article-title":"Compact modeling of RRAM devices and its applications in 1T1R and 1S1R array design","volume":"62","author":"Chen","year":"2015","journal-title":"IEEE Trans. Electron Devices"},{"key":"nceaceea9bib15","doi-asserted-by":"publisher","first-page":"1811","DOI":"10.1039\/D0NA00100G","article-title":"Research progress on solutions to the sneak path issue in memristor crossbar arrays","volume":"2","author":"Shi","year":"2020","journal-title":"Nanoscale Adv."},{"key":"nceaceea9bib16","doi-asserted-by":"publisher","first-page":"8475","DOI":"10.1038\/ncomms9475","article-title":"Wafer-scale growth of VO2 thin films using a combinatorial approach","volume":"6","author":"Zhang","year":"2015","journal-title":"Nat. Commun."},{"key":"nceaceea9bib17","doi-asserted-by":"publisher","first-page":"5019","DOI":"10.1038\/s41467-021-24740-2","article-title":"Heterogeneous integration of single-crystalline rutile nanomembranes with steep phase transition on silicon substrates","volume":"12","author":"Lee","year":"2021","journal-title":"Nat. Commun."},{"key":"nceaceea9bib18","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevB.91.134105","article-title":"Reconfigurable anisotropy and functional transformations with VO2-based metamaterial electric circuits","volume":"91","author":"Savo","year":"2015","journal-title":"Phys. Rev. B"},{"key":"nceaceea9bib19","doi-asserted-by":"publisher","first-page":"679","DOI":"10.1002\/adfm.201402687","article-title":"Metal\u2010insulator transition in ALD VO2 ultrathin films and nanoparticles: morphological control","volume":"25","author":"Peter","year":"2015","journal-title":"Adv. Funct. Mater."}],"container-title":["Neuromorphic Computing and Engineering"],"original-title":[],"link":[{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2634-4386\/aceea9","content-type":"text\/html","content-version":"am","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2634-4386\/aceea9\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2634-4386\/aceea9","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2634-4386\/aceea9\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2634-4386\/aceea9\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"syndication"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2634-4386\/aceea9\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2634-4386\/aceea9\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"similarity-checking"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2634-4386\/aceea9\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2023,9,25]],"date-time":"2023-09-25T09:13:58Z","timestamp":1695633238000},"score":1,"resource":{"primary":{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2634-4386\/aceea9"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2023,8,29]]},"references-count":19,"journal-issue":{"issue":"3","published-online":{"date-parts":[[2023,8,29]]},"published-print":{"date-parts":[[2023,9,1]]}},"URL":"https:\/\/doi.org\/10.1088\/2634-4386\/aceea9","relation":{},"ISSN":["2634-4386"],"issn-type":[{"value":"2634-4386","type":"electronic"}],"subject":[],"published":{"date-parts":[[2023,8,29]]},"assertion":[{"value":"Integration of Ag-CBRAM crossbars and Mott ReLU neurons for efficient implementation of deep neural networks in hardware","name":"article_title","label":"Article Title"},{"value":"Neuromorphic Computing and Engineering","name":"journal_title","label":"Journal Title"},{"value":"paper","name":"article_type","label":"Article Type"},{"value":"\u00a9 2023 The Author(s). Published by IOP Publishing Ltd","name":"copyright_information","label":"Copyright Information"},{"value":"2022-10-04","name":"date_received","label":"Date Received","group":{"name":"publication_dates","label":"Publication dates"}},{"value":"2023-08-09","name":"date_accepted","label":"Date Accepted","group":{"name":"publication_dates","label":"Publication dates"}},{"value":"2023-08-29","name":"date_epub","label":"Online publication date","group":{"name":"publication_dates","label":"Publication dates"}}]}}