{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,6,21]],"date-time":"2025-06-21T11:02:32Z","timestamp":1750503752113,"version":"3.37.3"},"reference-count":88,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","issue":"12","license":[{"start":{"date-parts":[[2020,12,1]],"date-time":"2020-12-01T00:00:00Z","timestamp":1606780800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2020,12,1]],"date-time":"2020-12-01T00:00:00Z","timestamp":1606780800000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2020,12,1]],"date-time":"2020-12-01T00:00:00Z","timestamp":1606780800000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"name":"Systems On Nanoscale Information fabriCs (SONIC) and the Center for Brain-Inspired Computing"},{"DOI":"10.13039\/100000028","name":"Semiconductor Research Corporation (SRC) and the Defense Advanced Research Projects Agency","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100000028","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Proc. IEEE"],"published-print":{"date-parts":[[2020,12]]},"DOI":"10.1109\/jproc.2020.3034117","type":"journal-article","created":{"date-parts":[[2020,11,9]],"date-time":"2020-11-09T21:07:04Z","timestamp":1604956024000},"page":"2251-2275","source":"Crossref","is-referenced-by-count":36,"title":["Deep In-Memory Architectures in SRAM: An Analog Approach to Approximate Computing"],"prefix":"10.1109","volume":"108","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-8104-5136","authenticated-orcid":false,"given":"Mingu","family":"Kang","sequence":"first","affiliation":[]},{"ORCID":"https:\/\/orcid.org\/0000-0003-4743-6461","authenticated-orcid":false,"given":"Sujan K.","family":"Gonugondla","sequence":"additional","affiliation":[]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4323-9164","authenticated-orcid":false,"given":"Naresh R.","family":"Shanbhag","sequence":"additional","affiliation":[]}],"member":"263","reference":[{"key":"ref73","first-page":"1737","article-title":"Deep learning with limited numerical precision","author":"gupta","year":"2015","journal-title":"Proc Int Conf Mach Learn (ICML)"},{"key":"ref72","doi-asserted-by":"publisher","DOI":"10.1109\/16.628833"},{"key":"ref71","doi-asserted-by":"publisher","DOI":"10.1145\/2775054.2694358"},{"key":"ref70","article-title":"A 481 pJ\/decision 3.4 M decision\/s multifunctional deep in-memory inference processor using standard 6T SRAM array","author":"kang","year":"2016","journal-title":"arXiv 1610 07501"},{"key":"ref76","doi-asserted-by":"publisher","DOI":"10.1007\/3-540-59119-2_166"},{"key":"ref77","doi-asserted-by":"publisher","DOI":"10.1007\/BF00058655"},{"key":"ref74","doi-asserted-by":"publisher","DOI":"10.1109\/ACSSC.2015.7421361"},{"key":"ref39","first-page":"234","article-title":"A 65 nm computing-in-memory-based CNN processor with 2.9-to-35.8 TOPS\/W system energy efficiency using dynamic-sparsity performance-scaling architecture and energy-efficient inter\/intra-macro data reuse","author":"yue","year":"2020","journal-title":"IEEE Int Solid-State Circuits Conf (ISSCC) Dig Tech Papers"},{"journal-title":"Center for Biologicaland Computationallearning (CBCL)","year":"2000","key":"ref75"},{"key":"ref38","doi-asserted-by":"publisher","DOI":"10.23919\/VLSIC.2019.8778028"},{"key":"ref78","doi-asserted-by":"publisher","DOI":"10.1023\/A:1010933404324"},{"key":"ref79","doi-asserted-by":"publisher","DOI":"10.1109\/ESSCIRC.2017.8094576"},{"key":"ref33","doi-asserted-by":"publisher","DOI":"10.1109\/VLSIT.2018.8510687"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.1109\/JSSC.2019.2899730"},{"key":"ref31","first-page":"496","article-title":"A 65 nm 4 Kb algorithm-dependent computing-in-memory SRAM unit-macro with 2.3 ns and 55.8TOPS\/W fully parallel product-sum operation for binary DNN edge processors","author":"khwa","year":"2018","journal-title":"IEEE Int Solid-State Circuits Conf (ISSCC) Dig Tech Papers"},{"key":"ref30","first-page":"494","article-title":"A 65 nm 1 Mb nonvolatile computing-in-memory ReRAM macro with sub-16ns multiply-and-accumulate for binary DNN AI edge processors","author":"chen","year":"2018","journal-title":"IEEE Int Solid-State Circuits Conf (ISSCC) Dig Tech Papers"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.23919\/VLSIC.2019.8778160"},{"key":"ref36","first-page":"248","article-title":"A ternary based bit scalable, 8.80 TOPS\/W CNN accelerator with many-core processing-in-memory architecture with 896K synapses\/mm2","author":"okumura","year":"2019","journal-title":"Proc IEEE Symp VLSI Technol"},{"key":"ref35","article-title":"A microprocessor implemented in 65 nm CMOS with configurable and bit-scalable accelerator for programmable in-memory computing","author":"jia","year":"2018","journal-title":"arXiv 1811 04047"},{"key":"ref34","doi-asserted-by":"publisher","DOI":"10.1109\/ISSCC.2019.8662392"},{"key":"ref60","first-page":"1","article-title":"A machine-learning classifier implemented in a standard 6T SRAM array","author":"zhang","year":"2016","journal-title":"Proc IEEE Symp VLSI Circuits (VLSI)"},{"key":"ref62","doi-asserted-by":"publisher","DOI":"10.1007\/BF00994018"},{"key":"ref61","article-title":"A microprocessor implemented in 65 nm CMOS with configurable and bit-scalable accelerator for programmable in-memory computing","author":"jia","year":"2018","journal-title":"arXiv 1811 04047"},{"key":"ref63","first-page":"21","article-title":"A theoretical framework for back-propagation","volume":"1","author":"lecun","year":"1988","journal-title":"Proc 1988 Connectionist Models Summer School"},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.1109\/JSSC.2018.2867275"},{"key":"ref64","doi-asserted-by":"publisher","DOI":"10.1109\/TCSI.2019.2960841"},{"key":"ref27","doi-asserted-by":"publisher","DOI":"10.1109\/JSSC.2018.2822703"},{"key":"ref65","doi-asserted-by":"publisher","DOI":"10.1109\/VLSIC.2018.8502421"},{"key":"ref66","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2015.7178127"},{"key":"ref29","doi-asserted-by":"publisher","DOI":"10.1109\/ISSCC.2018.8310397"},{"key":"ref67","doi-asserted-by":"publisher","DOI":"10.1109\/ASSCC.2016.7844125"},{"key":"ref68","doi-asserted-by":"publisher","DOI":"10.1109\/JPROC.2018.2869867"},{"key":"ref69","doi-asserted-by":"publisher","DOI":"10.1109\/ISCA.2018.00015"},{"key":"ref2","first-page":"1097","article-title":"ImageNet classification with deep convolutional neural networks","author":"krizhevsky","year":"2012","journal-title":"Proc Adv Neural Inf Process Syst (NIPS)"},{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.1038\/nature16961"},{"key":"ref20","first-page":"240","article-title":"A 28 nm 400 MHz 4-parallel 1.6 Gsearch\/s 80Mb ternary CAM","author":"nii","year":"2014","journal-title":"IEEE Int Solid-State Circuits Conf (ISSCC) Dig Tech Papers"},{"key":"ref22","first-page":"44","article-title":"A 3.14 $\\mu{\\mathrm{ m}}^{2}~4\\text{T}$\n-2MTJ-cell fully parallel TCAM based on nonvolatile logic-in-memory architecture","author":"matsunaga","year":"2012","journal-title":"Proc Symp VLSI circuits (VLSI)"},{"key":"ref21","first-page":"104c","article-title":"1Mb $0.41~\\mu$\nm2 2T-2R cell nonvolatile TCAM with two-bit encoding and clocked self-referenced sensing","author":"li","year":"2013","journal-title":"Proc Symp VLSI-Technol (VLSI-T)"},{"article-title":"Compute memory","year":"2017","author":"shanbhag","key":"ref24"},{"key":"ref23","first-page":"1","article-title":"ReRAM-based 4T2R nonvolatile TCAM with 7x NVM-stress reduction, and 4x improvement in speed-wordlength-capacity for normally-off instant-on filter-based search engines used in big-data processing","author":"huang","year":"2014","journal-title":"Proc Symp VLSI circuits (VLSI)"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1109\/JSSC.2017.2782087"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.1109\/JSSC.2016.2642198"},{"key":"ref50","doi-asserted-by":"publisher","DOI":"10.1109\/4.494201"},{"key":"ref51","doi-asserted-by":"publisher","DOI":"10.1145\/1837274.1837491"},{"key":"ref59","first-page":"490","article-title":"A 42pJ\/decision 3.12 TOPS\/W robust in-memory machine learning classifier with on-chip training","author":"gonugondla","year":"2018","journal-title":"IEEE Int Solid-State Circuits Conf (ISSCC) Dig Tech Papers"},{"key":"ref58","doi-asserted-by":"publisher","DOI":"10.1109\/TCAD.2012.2217962"},{"key":"ref57","first-page":"260","article-title":"A 21.5 M-query-vectors\/s 3.37nJ\/vector reconfigurable k-nearest-neighbor accelerator with adaptive precision in 14 nm tri-gate CMOS","author":"kaul","year":"2016","journal-title":"IEEE Int Solid-State Circuits Conf (ISSCC) Dig Tech Papers"},{"key":"ref56","first-page":"2074","article-title":"Learning structured sparsity in deep neural networks","author":"wen","year":"2016","journal-title":"Proc Adv Neural Inf Process Syst"},{"key":"ref55","doi-asserted-by":"publisher","DOI":"10.1109\/ICCAD.2013.6691202"},{"key":"ref54","doi-asserted-by":"publisher","DOI":"10.1109\/ICCAD.2014.7001398"},{"key":"ref53","doi-asserted-by":"publisher","DOI":"10.1109\/NANO.2013.6720793"},{"key":"ref52","doi-asserted-by":"publisher","DOI":"10.1145\/2465787.2465794"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.1109\/HPCA.2017.21"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1109\/ISSCC.2019.8662419"},{"key":"ref40","first-page":"240","article-title":"A 28 nm 64 Kb inference-training two-way transpose multibit 6T SRAM compute-in-memory macro for AI edge chips","author":"su","year":"2020","journal-title":"IEEE Int Solid-State Circuits Conf (ISSCC) Dig Tech Papers"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.1109\/JSSC.2018.2869150"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1109\/VLSIT.2012.6242474"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.1109\/IEDM.2014.7047120"},{"key":"ref15","doi-asserted-by":"crossref","first-page":"1009","DOI":"10.1109\/JSSC.2016.2515510","article-title":"A 28 nm configurable memory (TCAM\/BCAM\/SRAM) using push-rule 6T bit cell enabling logic-in-memory","volume":"51","author":"jeloka","year":"2016","journal-title":"IEEE J Solid-State Circuits"},{"key":"ref82","doi-asserted-by":"publisher","DOI":"10.1109\/FCCM.2012.47"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2014.6855225"},{"key":"ref81","doi-asserted-by":"publisher","DOI":"10.1109\/ICPR.2010.816"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-35971-3"},{"key":"ref84","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2018.8461702"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1109\/MSSC.2019.2922889"},{"key":"ref83","first-page":"3007","article-title":"Analytical guarantees on numerical precision of deep neural networks","author":"sakr","year":"2017","journal-title":"Proc Int Conf Mach Learn (ICML)"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.1109\/JSSC.2010.2082270"},{"key":"ref80","doi-asserted-by":"publisher","DOI":"10.1109\/4.210039"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.1109\/JSSC.2016.2616357"},{"key":"ref3","first-page":"10","article-title":"1.1 Computing’s energy problem (and what we can do about it)","author":"horowitz","year":"2014","journal-title":"IEEE Int Solid-State Circuits Conf (ISSCC) Dig Tech Papers"},{"key":"ref6","first-page":"246","article-title":"14.5 envision: A 0.26-to-10TOPS\/W subword-parallel dynamic-voltage-accuracy-frequency-scalable convolutional neural network processor in 28 nm FDSOI","author":"moons","year":"2017","journal-title":"IEEE Int Solid-State Circuits Conf (ISSCC) Dig Tech Papers"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1145\/2644865.2541967"},{"key":"ref85","first-page":"348","article-title":"Accurate and efficient 2-bit quantized neural networks","author":"choi","year":"2019","journal-title":"Proc SysML Conf"},{"key":"ref8","first-page":"242","article-title":"A 28 nm SoC with a 1.2 GHz 568 nJ\/prediction sparse deep-neural-network engine with >0.1 timing error rate tolerance for IoT applications","author":"whatmough","year":"2017","journal-title":"IEEE Int Solid-State Circuits Conf (ISSCC) Dig Tech Papers"},{"key":"ref86","doi-asserted-by":"publisher","DOI":"10.1162\/089976698300017052"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1109\/ISSCC.2017.7870350"},{"key":"ref49","doi-asserted-by":"publisher","DOI":"10.1023\/A:1007986707921"},{"key":"ref87","doi-asserted-by":"publisher","DOI":"10.1109\/JSSC.2015.2408332"},{"key":"ref88","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2019.8683521"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1109\/ISSCC.2018.8310262"},{"key":"ref46","first-page":"4107","article-title":"Binarized neural networks","author":"hubara","year":"2016","journal-title":"Proc Adv Neural Inf Process Syst"},{"key":"ref45","first-page":"7675","article-title":"Training deep neural networks with 8-bit floating point numbers","author":"wang","year":"2018","journal-title":"Proc Adv Neural Inf Process Syst"},{"key":"ref48","first-page":"145","article-title":"Multiplier-less artificial neurons exploiting error resiliency for energy-efficient neural computing","author":"sarwar","year":"2016","journal-title":"Proc Design Autom Test Eur Conf Exhibition (DATE)"},{"key":"ref47","doi-asserted-by":"publisher","DOI":"10.1145\/2966986.2967021"},{"key":"ref42","first-page":"246","article-title":"A 28 nm 64 Kb 6T SRAM computing-in-memory macro with 8b MAC operation for AI edge chips","author":"si","year":"2020","journal-title":"IEEE Int Solid-State Circuits Conf (ISSCC) Dig Tech Papers"},{"key":"ref41","first-page":"242","article-title":"A 351 TOPS\/W and 372.4 GOPS compute-in-memory SRAM macro in 7 nm FinFET CMOS for machine learning applications","author":"dong","year":"2020","journal-title":"IEEE Int Solid-State Circuits Conf (ISSCC) Dig Tech Papers"},{"key":"ref44","article-title":"Pruning filters for efficient ConvNets","author":"li","year":"2016","journal-title":"arXiv 1608 08710"},{"key":"ref43","first-page":"244","article-title":"A 22 nm 2 Mb ReRAM compute-in-memory macro with 121–28 TOPS\/W for multibit MAC computing for tiny AI edge devices","author":"xue","year":"2020","journal-title":"IEEE Int Solid-State Circuits Conf (ISSCC) Dig Tech Papers"}],"container-title":["Proceedings of the IEEE"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx7\/5\/9264766\/09252843.pdf?arnumber=9252843","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2022,4,27]],"date-time":"2022-04-27T13:40:21Z","timestamp":1651066821000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/9252843\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2020,12]]},"references-count":88,"journal-issue":{"issue":"12"},"URL":"https:\/\/doi.org\/10.1109\/jproc.2020.3034117","relation":{},"ISSN":["0018-9219","1558-2256"],"issn-type":[{"type":"print","value":"0018-9219"},{"type":"electronic","value":"1558-2256"}],"subject":[],"published":{"date-parts":[[2020,12]]}}}