{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,1]],"date-time":"2026-10-01T01:55:21Z","timestamp":1790819721484,"version":"4.1.0"},"reference-count":51,"publisher":"Association for Computing Machinery (ACM)","issue":"6","license":[{"start":{"date-parts":[[2022,10,18]],"date-time":"2022-10-18T00:00:00Z","timestamp":1666051200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"name":"ECSEL Joint Undertaking project TEMPO in collaboration with the European Union\u2019s H2020 Framework Program","award":["H2020\/2014-2020"],"award-info":[{"award-number":["H2020\/2014-2020"]}]},{"name":"National Authorities","award":["826655"],"award-info":[{"award-number":["826655"]}]},{"name":"Carl Zeiss Foundation under the grant Sustainable Embedded AI"}],"content-domain":{"domain":["dl.acm.org"],"crossmark-restriction":true},"short-container-title":["ACM Trans. Embed. Comput. Syst."],"published-print":{"date-parts":[[2022,11,30]]},"abstract":"<jats:p>\n            Today, a large number of applications depend on deep neural networks (DNN) to process data and perform complicated tasks at restricted power and latency specifications. Therefore, processing-in-memory (PIM) platforms are actively explored as a promising approach to improve the throughput and the energy efficiency of DNN computing systems. Several PIM architectures adopt resistive non-volatile memories as their main unit to build crossbar-based accelerators for DNN inference. However, these structures suffer from several drawbacks such as reliability, low accuracy, large ADCs\/DACs power consumption and area, high write energy, and so on. In this article, we present a new mixed-signal in-memory architecture based on the bit-decomposition of the multiply and accumulate (MAC) operations. Our in-memory inference architecture uses a single FeFET as a non-volatile memory cell. Compared to the prior work, this system architecture provides a high level of parallelism while using only 3-bit ADCs. Also, it eliminates the need for any DAC. In addition, we provide flexibility and a very high utilization efficiency even for varying tasks and loads. Simulations demonstrate that we outperform state-of-the-art efficiencies with 36.5 TOPS\/W and can pack 2.05 TOPS with 8-bit activation and 4-bit weight precision in an area of 4.9 mm\n            <jats:sup>2<\/jats:sup>\n            using 22 nm FDSOI technology. Employing binary operation, we obtain 1169 TOPS\/W and over 261 TOPS\/W\/mm\n            <jats:sup>2<\/jats:sup>\n            on system level.\n          <\/jats:p>","DOI":"10.1145\/3529760","type":"journal-article","created":{"date-parts":[[2022,4,9]],"date-time":"2022-04-09T09:07:01Z","timestamp":1649495221000},"page":"1-25","update-policy":"https:\/\/doi.org\/10.1145\/crossmark-policy","source":"Crossref","is-referenced-by-count":39,"title":["FELIX: A Ferroelectric FET Based Low Power Mixed-Signal In-Memory Architecture for DNN Acceleration"],"prefix":"10.1145","volume":"21","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-9421-9489","authenticated-orcid":false,"given":"Taha","family":"Soliman","sequence":"first","affiliation":[{"name":"Robert Bosch GmbH, Renningen, Germany"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-2445-9989","authenticated-orcid":false,"given":"Nellie","family":"Laleni","sequence":"additional","affiliation":[{"name":"Fraunhofer IPMS, Center Nanoelectronic Technologies, Dresden, Germany"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-4648-8681","authenticated-orcid":false,"given":"Tobias","family":"Kirchner","sequence":"additional","affiliation":[{"name":"Robert Bosch GmbH, Renningen, Germany"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-6564-9121","authenticated-orcid":false,"given":"Franz","family":"M\u00fcller","sequence":"additional","affiliation":[{"name":"Fraunhofer IPMS, Center Nanoelectronic Technologies, Dresden, Germany"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-3278-7392","authenticated-orcid":false,"given":"Ashish","family":"Shrivastava","sequence":"additional","affiliation":[{"name":"Fraunhofer IPMS, Center Nanoelectronic Technologies, Dresden, Germany"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4672-8676","authenticated-orcid":false,"given":"Thomas","family":"K\u00e4mpfe","sequence":"additional","affiliation":[{"name":"Fraunhofer IPMS, Center Nanoelectronic Technologies, Dresden, Germany"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-4144-0283","authenticated-orcid":false,"given":"Andre","family":"Guntoro","sequence":"additional","affiliation":[{"name":"Robert Bosch GmbH, Renningen, Germany"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-9010-086X","authenticated-orcid":false,"given":"Norbert","family":"Wehn","sequence":"additional","affiliation":[{"name":"University of Kaiserslautern, Kaiserslautern, Germany"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"320","published-online":{"date-parts":[[2022,10,18]]},"reference":[{"key":"e_1_3_1_2_2","doi-asserted-by":"publisher","DOI":"10.1109\/TED.2020.2995781"},{"key":"e_1_3_1_3_2","doi-asserted-by":"publisher","DOI":"10.1038\/s41586-018-0180-5"},{"key":"e_1_3_1_4_2","doi-asserted-by":"publisher","DOI":"10.1109\/JSSC.2017.2778702"},{"key":"e_1_3_1_5_2","doi-asserted-by":"publisher","DOI":"10.1145\/3240765.3240799"},{"key":"e_1_3_1_6_2","doi-asserted-by":"publisher","DOI":"10.1109\/IMW48823.2020.9108150"},{"key":"e_1_3_1_7_2","doi-asserted-by":"publisher","DOI":"10.1063\/1.3634052"},{"key":"e_1_3_1_8_2","doi-asserted-by":"publisher","DOI":"10.1109\/TCAD.2019.2917852"},{"key":"e_1_3_1_9_2","doi-asserted-by":"publisher","DOI":"10.23919\/DATE.2018.8342199"},{"key":"e_1_3_1_10_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2016.350"},{"key":"e_1_3_1_11_2","doi-asserted-by":"publisher","DOI":"10.1109\/IEDM19573.2019.8993599"},{"key":"e_1_3_1_12_2","doi-asserted-by":"publisher","DOI":"10.23919\/DATE48585.2020.9116476"},{"key":"e_1_3_1_13_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2009.5206848"},{"key":"e_1_3_1_14_2","doi-asserted-by":"publisher","DOI":"10.1109\/MSP.2012.2211477"},{"key":"e_1_3_1_15_2","doi-asserted-by":"publisher","DOI":"10.1109\/IEDM.2017.8268425"},{"key":"e_1_3_1_16_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2016.90"},{"key":"e_1_3_1_17_2","unstructured":"Forrest N. Iandola Song Han Matthew W. Moskewicz Khalid Ashraf William J. Dally and Kurt Keutzer. 2016. SqueezeNet: AlexNet-level accuracy with 50\u00d7 fewer parameters and <0.5MB model size. arXiv:1602.07360. Retrieved from https:\/\/arxiv.org\/abs\/1602.07360."},{"key":"e_1_3_1_18_2","unstructured":"Sambhav R. Jain Albert Gural Michael Wu and Chris Dick. 2019. Trained Quantization Thresholds for Accurate and Efficient Fixed-Point Inference of Deep Neural Networks. arXiv:1903.08066. Retrieved from https:\/\/arxiv.org\/abs\/1903.08066."},{"key":"e_1_3_1_19_2","doi-asserted-by":"publisher","DOI":"10.1109\/LCA.2016.2597140"},{"key":"e_1_3_1_20_2","doi-asserted-by":"publisher","DOI":"10.1109\/MM.2020.3026667"},{"key":"e_1_3_1_21_2","doi-asserted-by":"publisher","DOI":"10.1109\/ESSCIRC.2019.8902824"},{"key":"e_1_3_1_22_2","volume-title":"CIFAR-10 (Canadian Institute for Advanced Research)","author":"Krizhevsky Alex","year":"2009","unstructured":"Alex Krizhevsky, Vinod Nair, and Geoffrey Hinton. 2009. CIFAR-10 (Canadian Institute for Advanced Research). Retrieved June 2021 from http:\/\/www.cs.toronto.edu\/kriz\/cifar.html."},{"key":"e_1_3_1_23_2","doi-asserted-by":"publisher","DOI":"10.1109\/MICRO.2018.00021"},{"key":"e_1_3_1_24_2","doi-asserted-by":"publisher","DOI":"10.1109\/5.726791"},{"key":"e_1_3_1_25_2","doi-asserted-by":"publisher","DOI":"10.1063\/1.5129318"},{"key":"e_1_3_1_26_2","doi-asserted-by":"publisher","DOI":"10.1109\/TED.2021.3068716"},{"key":"e_1_3_1_27_2","doi-asserted-by":"publisher","DOI":"10.1109\/LED.2020.3031308"},{"key":"e_1_3_1_28_2","doi-asserted-by":"publisher","DOI":"10.1109\/ISSCC.2018.8310262"},{"key":"e_1_3_1_29_2","doi-asserted-by":"publisher","DOI":"10.1145\/2744769.2744870"},{"key":"e_1_3_1_30_2","doi-asserted-by":"publisher","DOI":"10.1109\/JXCDC.2019.2923745"},{"key":"e_1_3_1_31_2","doi-asserted-by":"publisher","DOI":"10.1109\/TVLSI.2018.2819190"},{"key":"e_1_3_1_32_2","doi-asserted-by":"publisher","DOI":"10.1109\/LED.2002.1015207"},{"key":"e_1_3_1_33_2","doi-asserted-by":"publisher","DOI":"10.1063\/1.351910"},{"key":"e_1_3_1_34_2","doi-asserted-by":"publisher","DOI":"10.1149\/2.0081505jss"},{"key":"e_1_3_1_35_2","doi-asserted-by":"publisher","DOI":"10.1109\/JEDS.2018.2817628"},{"key":"e_1_3_1_36_2","unstructured":"IEEE Conference on Computer Vision and Pattern Recognition 2018 Inverted residuals and linear bottlenecks: Mobile networks for classification detection and segmentation"},{"key":"e_1_3_1_37_2","doi-asserted-by":"publisher","DOI":"10.1109\/ISCAS51556.2021.9401196"},{"key":"e_1_3_1_38_2","doi-asserted-by":"publisher","DOI":"10.1145\/3007787.3001139"},{"key":"e_1_3_1_39_2","doi-asserted-by":"publisher","DOI":"10.1109\/IEDM13553.2020.9372124"},{"key":"e_1_3_1_40_2","doi-asserted-by":"publisher","DOI":"10.1109\/SOCC49529.2020.9524750"},{"key":"e_1_3_1_41_2","doi-asserted-by":"publisher","DOI":"10.1109\/ASAP49362.2020.00027"},{"key":"e_1_3_1_42_2","doi-asserted-by":"publisher","DOI":"10.1109\/HPCA.2017.55"},{"key":"e_1_3_1_43_2","doi-asserted-by":"publisher","DOI":"10.1109\/LED.2018.2791510"},{"key":"e_1_3_1_44_2","doi-asserted-by":"publisher","DOI":"10.1109\/JPROC.2017.2684830"},{"key":"e_1_3_1_45_2","doi-asserted-by":"publisher","DOI":"10.1109\/TPDS.2019.2937517"},{"key":"e_1_3_1_46_2","unstructured":"Fisher Yu and Vladlen Koltun. 2015. Multi-Scale Context Aggregation by Dilated Convolutions. arXiv:1511.07122. Retrieved from https:\/\/arxiv.org\/abs\/1511.07122."},{"key":"e_1_3_1_47_2","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-54313-0"},{"key":"e_1_3_1_48_2","doi-asserted-by":"publisher","DOI":"10.1109\/JPROC.2018.2790840"},{"key":"e_1_3_1_49_2","doi-asserted-by":"publisher","DOI":"10.1109\/IEDM.2016.7838429"},{"key":"e_1_3_1_50_2","doi-asserted-by":"publisher","DOI":"10.1109\/ACCESS.2020.3000865"},{"key":"e_1_3_1_51_2","doi-asserted-by":"publisher","DOI":"10.1109\/DAC18072.2020.9218590"},{"key":"e_1_3_1_52_2","doi-asserted-by":"publisher","DOI":"10.1145\/3316781.3317739"}],"container-title":["ACM Transactions on Embedded Computing Systems"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3529760","content-type":"unspecified","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/dl.acm.org\/doi\/pdf\/10.1145\/3529760","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,6,17]],"date-time":"2025-06-17T19:31:24Z","timestamp":1750188684000},"score":1,"resource":{"primary":{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3529760"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2022,10,18]]},"references-count":51,"journal-issue":{"issue":"6","published-print":{"date-parts":[[2022,11,30]]}},"alternative-id":["10.1145\/3529760"],"URL":"https:\/\/doi.org\/10.1145\/3529760","relation":{},"ISSN":["1539-9087","1558-3465"],"issn-type":[{"value":"1539-9087","type":"print"},{"value":"1558-3465","type":"electronic"}],"subject":[],"published":{"date-parts":[[2022,10,18]]},"assertion":[{"value":"2021-07-12","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2022-03-30","order":1,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2022-10-18","order":2,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}