{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,24]],"date-time":"2026-09-24T19:59:28Z","timestamp":1790279968700,"version":"4.1.0"},"reference-count":219,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","license":[{"start":{"date-parts":[[2026,1,1]],"date-time":"2026-01-01T00:00:00Z","timestamp":1767225600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/legalcode"},{"start":{"date-parts":[[2026,1,1]],"date-time":"2026-01-01T00:00:00Z","timestamp":1767225600000},"content-version":"am","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/legalcode"}],"funder":[{"name":"National Science Foundation","award":["2233893"],"award-info":[{"award-number":["2233893"]}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE Access"],"published-print":{"date-parts":[[2026]]},"DOI":"10.1109\/access.2026.3734236","type":"journal-article","created":{"date-parts":[[2026,9,23]],"date-time":"2026-09-23T19:19:58Z","timestamp":1790191198000},"page":"144280-144333","source":"Crossref","is-referenced-by-count":0,"title":["A Cross-Layer Survey of Low-Power Machine Learning From Algorithm to Hardware"],"prefix":"10.1109","volume":"14","author":[{"ORCID":"https:\/\/orcid.org\/0009-0007-9581-8348","authenticated-orcid":false,"given":"Zuxiong","family":"Tan","sequence":"first","affiliation":[{"name":"University of California, Davis","place":["Davis, CA, USA"],"department":["Department of Electrical and Computer Engineering"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-6075-0880","authenticated-orcid":false,"given":"Ali","family":"Karkehabadi","sequence":"additional","affiliation":[{"name":"University of California, Davis","place":["Davis, CA, USA"],"department":["Department of Electrical and Computer Engineering"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ning","family":"Miao","sequence":"additional","affiliation":[{"name":"University of California, Davis","place":["Davis, CA, USA"],"department":["Department of Electrical and Computer Engineering"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0004-6763-0290","authenticated-orcid":false,"given":"Wei","family":"Shao","sequence":"additional","affiliation":[{"name":"University of California, Davis","place":["Davis, CA, USA"],"department":["Department of Computer Science"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-8904-4699","authenticated-orcid":false,"given":"Houman","family":"Homayoun","sequence":"additional","affiliation":[{"name":"University of California, Davis","place":["Davis, CA, USA"],"department":["Department of Electrical and Computer Engineering"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4052-8075","authenticated-orcid":false,"given":"Avesta","family":"Sasan","sequence":"additional","affiliation":[{"name":"University of California, Davis","place":["Davis, CA, USA"],"department":["Department of Electrical and Computer Engineering"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/p19-1355"},{"key":"ref2","article-title":"Carbon emissions and large neural network training","author":"Patterson","year":"2021","journal-title":"arXiv:2104.10350"},{"key":"ref3","article-title":"2024 United States data center energy usage report","author":"Shehabi","year":"2024"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.1145\/3381831"},{"issue":"248","key":"ref5","first-page":"1","article-title":"Towards the systematic reporting of the energy and carbon footprints of machine learning","volume":"21","author":"Henderson","year":"2020","journal-title":"J. Mach. Learn. Res."},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.1109\/hpca61900.2025.00092"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1002\/widm.1507"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.3390\/en18112810"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1109\/ISSCC.2014.6757323"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.1109\/IPDPS.2013.77"},{"key":"ref11","first-page":"119","article-title":"Zeus: Understanding and optimizing GPU energy consumption of DNN training","volume-title":"Proc. 20th USENIX Symp. Networked Syst. Design Implement. (NSDI)","author":"You"},{"key":"ref12","article-title":"A survey on green deep learning","author":"Xu","year":"2021","journal-title":"arXiv:2111.05193"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1109\/JPROC.2017.2761740"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.1109\/ISCA.2016.40"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.1016\/j.jnca.2025.104375"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.37547\/tajet\/v7i11-310"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1016\/j.sysarc.2025.103501"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1109\/TNNLS.2024.3430028"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.3390\/electronics14122495"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.3390\/s25030846"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.1016\/j.rineng.2025.107849"},{"key":"ref22","article-title":"Energy-aware deep learning on resource-constrained hardware","author":"Millar","year":"2025","journal-title":"arXiv:2505.12523"},{"key":"ref23","doi-asserted-by":"publisher","DOI":"10.30564\/jeis.v7i2.10982"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.1109\/TNNLS.2024.3394494"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2025.acl-long.1563"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1109\/JPROC.2024.3437365"},{"key":"ref27","doi-asserted-by":"publisher","DOI":"10.1088\/2515-7620\/acf81b"},{"key":"ref28","article-title":"MLPerf tiny benchmark","volume-title":"Proc. NeurIPS Track Datasets Benchmarks","author":"Banbury"},{"key":"ref29","doi-asserted-by":"publisher","DOI":"10.1145\/3177754"},{"key":"ref30","article-title":"Codecarbon: Estimate and track carbon emissions from machine learning computing","author":"Schmidt","year":"2021"},{"key":"ref31","volume-title":"NVML API Reference Manual","year":"2025"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.1109\/SC41406.2024.00028"},{"key":"ref33","doi-asserted-by":"publisher","DOI":"10.1109\/4.535411"},{"key":"ref34","doi-asserted-by":"publisher","DOI":"10.52202\/085713-3652"},{"key":"ref35","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2017.643"},{"key":"ref36","article-title":"Deep compression: Compressing deep neural networks with pruning, trained quantization and Huffman coding","volume-title":"Proc. Int. Conf. Learn. Represent. (ICLR)","author":"Han"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.1145\/3079856.3080254"},{"key":"ref38","doi-asserted-by":"publisher","DOI":"10.1109\/HPCA56546.2023.10071080"},{"key":"ref39","doi-asserted-by":"publisher","DOI":"10.1145\/3466752.3480043"},{"key":"ref40","article-title":"Compression of deep convolutional neural networks for fast and low power mobile applications","volume-title":"Proc. Int. Conf. Learn. Represent. (ICLR)","author":"Kim"},{"key":"ref41","doi-asserted-by":"publisher","DOI":"10.1109\/IISWC63097.2024.00026"},{"key":"ref42","first-page":"264","article-title":"TIE: Energy-efficient tensor train-based inference engine for deep neural network","volume-title":"Proc. 46th Int. Symp. Comput. Archit.","author":"Deng"},{"key":"ref43","doi-asserted-by":"publisher","DOI":"10.1145\/3579371.3589103"},{"key":"ref44","first-page":"129","article-title":"The impact of knowledge distillation on the energy consumption and runtime efficiency of NLP models","volume-title":"Proc. IEEE\/ACM 3rd Int. Conf. AI Eng. Softw. Eng. AI","author":"Yuan"},{"key":"ref45","doi-asserted-by":"publisher","DOI":"10.1145\/3746027.3755609"},{"key":"ref46","doi-asserted-by":"publisher","DOI":"10.1109\/IJCNN64981.2025.11228858"},{"key":"ref47","doi-asserted-by":"publisher","DOI":"10.52202\/085713-4859"},{"key":"ref48","doi-asserted-by":"publisher","DOI":"10.1145\/3597503.3639156"},{"key":"ref49","doi-asserted-by":"publisher","DOI":"10.1109\/TCAD.2024.3519055"},{"key":"ref50","doi-asserted-by":"publisher","DOI":"10.23919\/DATE56975.2023.10137095"},{"key":"ref51","article-title":"Channel gating neural networks","volume-title":"Proc. Adv. Neural Inf. Process. Syst. (NeurIPS)","author":"Hua"},{"key":"ref52","doi-asserted-by":"publisher","DOI":"10.1145\/3510003.3510088"},{"key":"ref53","doi-asserted-by":"publisher","DOI":"10.1016\/j.future.2024.107606"},{"key":"ref54","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v37i7.26042"},{"key":"ref55","doi-asserted-by":"publisher","DOI":"10.1145\/3615991.3616405"},{"key":"ref56","doi-asserted-by":"publisher","DOI":"10.1145\/3634769.3634809"},{"key":"ref57","doi-asserted-by":"publisher","DOI":"10.1145\/3725843.3756038"},{"key":"ref58","first-page":"18332","article-title":"Deepspeed-MoE: Advancing mixture-of-experts inference and training to power next-generation AI at scale","volume-title":"Proc. 39th Int. Conf. Mach. Learn. (ICML)","volume":"162","author":"Rajbhandari"},{"key":"ref59","doi-asserted-by":"publisher","DOI":"10.1145\/3727200.3727217"},{"key":"ref60","doi-asserted-by":"publisher","DOI":"10.1109\/LCA.2025.3592563"},{"key":"ref61","doi-asserted-by":"publisher","DOI":"10.1109\/IPDPS65963.2026.00069"},{"key":"ref62","first-page":"1135","article-title":"Learning both weights and connections for efficient neural networks","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"28","author":"Han"},{"key":"ref63","article-title":"Accelerating sparse deep neural networks","author":"Mishra","year":"2021","journal-title":"arXiv:2104.08378"},{"key":"ref64","first-page":"598","article-title":"Optimal brain damage","volume-title":"Proc. Adv. Neural Inf. Process. Syst. (NeurIPS)","volume":"2","author":"LeCun"},{"key":"ref65","first-page":"164","article-title":"Second order derivatives for network pruning: Optimal brain surgeon","volume-title":"Proc. Adv. Neural Inf. Process. Syst. (NeurIPS)","author":"Hassibi"},{"key":"ref66","doi-asserted-by":"publisher","DOI":"10.52202\/079017-1063"},{"key":"ref67","first-page":"1269","article-title":"Exploiting linear structure within convolutional networks for efficient evaluation","volume-title":"Proc. Adv. Neural Inf. Process. Syst. (NeurIPS)","author":"Denton"},{"key":"ref68","doi-asserted-by":"publisher","DOI":"10.5244\/C.28.88"},{"key":"ref69","article-title":"Speeding-up convolutional neural networks using fine-tuned CP-decomposition","volume-title":"Proc. Int. Conf. Learn. Represent. (ICLR) Workshop","author":"Lebedev"},{"key":"ref70","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1503.02531"},{"key":"ref71","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2020.findings-emnlp.372"},{"key":"ref72","doi-asserted-by":"publisher","DOI":"10.52202\/079017-0040"},{"key":"ref73","doi-asserted-by":"publisher","DOI":"10.1145\/3694715.3695964"},{"key":"ref74","doi-asserted-by":"publisher","DOI":"10.1109\/IWQoS61813.2024.10682917"},{"key":"ref75","first-page":"2285","article-title":"Compressing neural networks with the hashing trick","volume-title":"Proc. Int. Conf. Mach. Learn. (ICML)","author":"Chen"},{"key":"ref76","doi-asserted-by":"publisher","DOI":"10.1145\/2939672.2939839"},{"key":"ref77","doi-asserted-by":"publisher","DOI":"10.1109\/SLT48900.2021.9383575"},{"key":"ref78","doi-asserted-by":"publisher","DOI":"10.1016\/j.suscom.2023.100857"},{"key":"ref79","doi-asserted-by":"publisher","DOI":"10.1109\/ICPR.2016.7900006"},{"key":"ref80","first-page":"3301","article-title":"Shallow-deep networks: Understanding and mitigating network overthinking","volume-title":"Proc. 36th Int. Conf. Mach. Learn. (ICML)","author":"Kaya"},{"key":"ref81","doi-asserted-by":"publisher","DOI":"10.1016\/j.neunet.2023.10.003"},{"key":"ref82","doi-asserted-by":"publisher","DOI":"10.1109\/DAC56929.2023.10247701"},{"key":"ref83","article-title":"Estimating or propagating gradients through stochastic neurons for conditional computation","author":"Bengio","year":"2013","journal-title":"arXiv:1308.3432"},{"key":"ref84","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2018.00919"},{"key":"ref85","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-01261-8_25"},{"key":"ref86","doi-asserted-by":"publisher","DOI":"10.1145\/3551349.3561158"},{"key":"ref87","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR42600.2020.00239"},{"key":"ref88","article-title":"Slimmable neural networks","volume-title":"Proc. Int. Conf. Learn. Represent. (ICLR)","author":"Yu"},{"key":"ref89","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2019.00189"},{"key":"ref90","article-title":"Once-for-all: Train one network and specialize it for efficient deployment","volume-title":"Proc. Int. Conf. Learn. Represent. (ICLR)","author":"Cai"},{"key":"ref91","doi-asserted-by":"publisher","DOI":"10.1145\/3689632"},{"key":"ref92","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2022.3194044"},{"key":"ref93","doi-asserted-by":"publisher","DOI":"10.1093\/nsr\/nwae088"},{"key":"ref94","doi-asserted-by":"publisher","DOI":"10.1109\/EDGE60047.2023.00034"},{"key":"ref95","doi-asserted-by":"publisher","DOI":"10.1162\/neco.1991.3.1.79"},{"key":"ref96","article-title":"Outrageously large neural networks: The sparsely-gated mixture-of-experts layer","volume-title":"Proc. Int. Conf. Learn. Represent. (ICLR)","author":"Shazeer"},{"issue":"120","key":"ref97","first-page":"1","article-title":"Switch transformers: Scaling to trillion parameter models with simple and efficient sparsity","volume":"23","author":"Fedus","year":"2021","journal-title":"J. Mach. Learn. Res."},{"key":"ref98","article-title":"SiDA-MoE: Sparsity-inspired data-aware serving for efficient and scalable large mixture-of-experts models","volume-title":"Proc. 7th Conf. Mach. Learn. Syst. (MLSys)","author":"Du"},{"key":"ref99","doi-asserted-by":"publisher","DOI":"10.3390\/electronics14030638"},{"key":"ref100","article-title":"VS-quant: Per-vector scaled quantization for accurate low-precision neural network inference","volume-title":"Proc. Mach. Learn. Syst.","author":"Dai"},{"key":"ref101","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-46493-0_32"},{"key":"ref102","article-title":"Mixed precision training","volume-title":"Proc. Int. Conf. Learn. Represent. (ICLR) Poster","author":"Micikevicius"},{"key":"ref103","article-title":"Campo: Cost-aware performance optimization for mixed-precision neural network training","volume-title":"Proc. USENIX Annu. Tech. Conf. (ATC)","author":"He"},{"key":"ref104","article-title":"FP8 formats for deep learning","author":"Micikevicius","year":"2022","journal-title":"arXiv:2209.05433"},{"key":"ref105","article-title":"Bridging the gap between promise and performance for microscaling FP4 quantization","volume-title":"Proc. Int. Conf. Learn. Represent. (ICLR)","author":"Egiazarian"},{"key":"ref106","article-title":"Execution-centric characterization of FP8 matrix cores, asynchronous execution, and structured sparsity on AMD MI300A","author":"Jarmusch","year":"2026","journal-title":"arXiv:2602.10262"},{"key":"ref107","doi-asserted-by":"publisher","DOI":"10.1145\/3712285.3759768"},{"key":"ref108","volume-title":"Introducing AMD CDNA 4 Architecture: Breakthrough AI and HPC Acceleration With Enhanced AI Capabilities, Advanced Precisions, and High Efficiency","year":"2025"},{"key":"ref109","article-title":"INT v.s. FP: A comprehensive study of fine-grained low-bit quantization formats","author":"Chen","year":"2025","journal-title":"arXiv:2510.25602"},{"key":"ref110","article-title":"FP8 versus INT8 for efficient deep learning inference","author":"van Baalen","year":"2023","journal-title":"arXiv:2303.17951"},{"key":"ref111","doi-asserted-by":"publisher","DOI":"10.1109\/ICSA-C63560.2024.00049"},{"key":"ref112","article-title":"Pretraining large language models with NVFP4","year":"2025","journal-title":"arXiv:2509.25149"},{"key":"ref113","first-page":"1737","article-title":"Deep learning with limited numerical precision","volume-title":"Proc. 32nd Int. Conf. Mach. Learn. (ICML)","author":"Gupta"},{"key":"ref114","first-page":"3123","article-title":"BinaryConnect: Training deep neural networks with binary weights during propagations","volume-title":"Proc. Adv. Neural Inf. Process. Syst. (NeurIPS)","author":"Courbariaux"},{"key":"ref115","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2018.00286"},{"key":"ref116","doi-asserted-by":"publisher","DOI":"10.1145\/3079856.3080246"},{"key":"ref117","article-title":"A study of BFLOAT16 for deep learning training","author":"Kalamkar","year":"2019","journal-title":"arXiv:1905.12322"},{"key":"ref118","article-title":"Microscaling data formats for deep learning","author":"Rouhani","year":"2023","journal-title":"arXiv:2310.10537"},{"key":"ref119","article-title":"Recipes for pre-training LLMs with MXFP8","author":"Mishra","year":"2025","journal-title":"arXiv:2506.08027"},{"key":"ref120","doi-asserted-by":"publisher","DOI":"10.52202\/079017-2193"},{"key":"ref121","article-title":"Repeated random sampling for minimizing the time-to-accuracy of learning","volume-title":"Proc. Int. Conf. Learn. Represent. (ICLR)","author":"Okanovic"},{"key":"ref122","doi-asserted-by":"publisher","DOI":"10.1007\/s10994-025-06907-w"},{"key":"ref123","article-title":"Dataset condensation with gradient matching","volume-title":"Proc. Int. Conf. Learn. Represent. (ICLR)","author":"Zhao"},{"key":"ref124","article-title":"Dataset distillation","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"31","author":"Wang"},{"key":"ref125","doi-asserted-by":"publisher","DOI":"10.1109\/GREENS66463.2025.00014"},{"key":"ref126","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.findings-eacl.143"},{"key":"ref127","article-title":"Dynamic resolution network","volume-title":"Proc. Adv. Neural Inf. Process. Syst. (NeurIPS)","author":"Zhu"},{"key":"ref128","doi-asserted-by":"publisher","DOI":"10.1109\/IJCNN64981.2025.11229265"},{"key":"ref129","first-page":"3690","article-title":"PoWER-BERT: Accelerating BERT inference via progressive word-vector elimination","volume-title":"Proc. 37th Int. Conf. Mach. Learn. (ICML)","volume":"1","author":"Goyal"},{"key":"ref130","doi-asserted-by":"publisher","DOI":"10.1145\/3534678.3539260"},{"key":"ref131","article-title":"DynamicViT: Efficient vision transformers with dynamic token sparsification","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Rao"},{"key":"ref132","doi-asserted-by":"publisher","DOI":"10.1109\/HPCA56546.2023.10071047"},{"key":"ref133","article-title":"Learning multiple layers of features from tiny images","author":"Krizhevsky","year":"2009"},{"key":"ref134","article-title":"Active learning for convolutional neural networks: A core-set approach","volume-title":"Proc. Int. Conf. Learn. Represent. (ICLR)","author":"Sener"},{"key":"ref135","article-title":"Deep learning on a data diet: Finding important examples early in training","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"34","author":"Paul"},{"key":"ref136","first-page":"6950","article-title":"Coresets for data-efficient training of machine learning models","volume-title":"Proc. 37th Int. Conf. Mach. Learn. (ICML)","volume":"1","author":"Mirzasoleiman"},{"key":"ref137","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v35i9.16988"},{"key":"ref138","doi-asserted-by":"publisher","DOI":"10.24963\/ijcai.2019\/372"},{"key":"ref139","doi-asserted-by":"publisher","DOI":"10.1109\/CVPRW56347.2022.00521"},{"key":"ref140","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.00364"},{"key":"ref141","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51701.2025.00056"},{"key":"ref142","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2020.emnlp-main.346"},{"key":"ref143","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2021.acllong.353"},{"key":"ref144","doi-asserted-by":"publisher","DOI":"10.52202\/075280-0848"},{"key":"ref145","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2016.91"},{"key":"ref146","article-title":"EfficientNet: Rethinking model scaling for convolutional neural networks","volume-title":"Proc. 36th Int. Conf. Mach. Learn. (ICML)","author":"Tan"},{"key":"ref147","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2021.acl-long.508"},{"key":"ref148","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00881"},{"key":"ref149","first-page":"578","article-title":"TVM: An automated end-to-end optimizing compiler for deep learning","volume-title":"Proc. 13th USENIX Symp. Operating Syst. Design Implement. (OSDI)","author":"Chen"},{"key":"ref150","doi-asserted-by":"publisher","DOI":"10.1145\/3520142"},{"key":"ref151","doi-asserted-by":"publisher","DOI":"10.1109\/JETCAS.2019.2910232"},{"key":"ref152","doi-asserted-by":"publisher","DOI":"10.1145\/3721288"},{"key":"ref153","first-page":"851","article-title":"EnvPipe: Performance-preserving DNN training framework for saving energy","volume-title":"Proc. USENIX Annu. Tech. Conf. (USENIX ATC)","author":"Choi"},{"key":"ref154","doi-asserted-by":"publisher","DOI":"10.1145\/3694715.3695970"},{"key":"ref155","doi-asserted-by":"publisher","DOI":"10.1109\/HPCA61900.2025.00102"},{"key":"ref156","article-title":"PolyThrottle: Energy-efficient neural network inference on edge devices","author":"Yan","year":"2023","journal-title":"arXiv:2310.19991"},{"key":"ref157","article-title":"Camel: Energy-aware LLM inference on resource-constrained devices","author":"Xu","year":"2025","journal-title":"arXiv:2508.09173"},{"key":"ref158","doi-asserted-by":"publisher","DOI":"10.1145\/3731569.3764808"},{"key":"ref159","doi-asserted-by":"publisher","DOI":"10.3390\/systems13090797"},{"key":"ref160","doi-asserted-by":"publisher","DOI":"10.1145\/3037697.3037698"},{"key":"ref161","doi-asserted-by":"publisher","DOI":"10.1109\/TMC.2022.3183098"},{"key":"ref162","doi-asserted-by":"publisher","DOI":"10.1145\/2307636.2307658"},{"key":"ref163","doi-asserted-by":"publisher","DOI":"10.1145\/1644893.1644927"},{"key":"ref164","doi-asserted-by":"publisher","DOI":"10.1109\/TWC.2020.3037554"},{"key":"ref165","doi-asserted-by":"publisher","DOI":"10.1109\/TWC.2023.3289177"},{"key":"ref166","doi-asserted-by":"publisher","DOI":"10.25080\/Majora-92bf1922-003"},{"key":"ref167","doi-asserted-by":"publisher","DOI":"10.1145\/2503210.2503296"},{"key":"ref168","doi-asserted-by":"publisher","DOI":"10.1145\/2491956.2462176"},{"key":"ref169","doi-asserted-by":"publisher","DOI":"10.1145\/113445.113449"},{"key":"ref170","doi-asserted-by":"publisher","DOI":"10.1007\/978-0-585-29603-6_17"},{"key":"ref171","doi-asserted-by":"publisher","DOI":"10.1109\/SFCS.1995.492493"},{"key":"ref172","doi-asserted-by":"publisher","DOI":"10.48550\/arxiv.1811.06965"},{"key":"ref173","doi-asserted-by":"publisher","DOI":"10.1145\/3341301.3359646"},{"key":"ref174","doi-asserted-by":"publisher","DOI":"10.1109\/ISCA.2018.00012"},{"key":"ref175","doi-asserted-by":"publisher","DOI":"10.1109\/TWC.2019.2946140"},{"key":"ref176","volume-title":"big.LITTLE Technology: The Future of Mobile","year":"2013"},{"key":"ref177","doi-asserted-by":"publisher","DOI":"10.1109\/ICCVW.2019.00447"},{"key":"ref178","doi-asserted-by":"publisher","DOI":"10.1145\/3229556.3229562"},{"key":"ref179","doi-asserted-by":"publisher","DOI":"10.1145\/3372224.3419194"},{"key":"ref180","doi-asserted-by":"publisher","DOI":"10.1145\/3450268.3453520"},{"key":"ref181","doi-asserted-by":"publisher","DOI":"10.1109\/TWC.2020.3024629"},{"key":"ref182","doi-asserted-by":"publisher","DOI":"10.1109\/ISCA.2016.41"},{"key":"ref183","doi-asserted-by":"publisher","DOI":"10.1145\/3620665.3640422"},{"key":"ref184","doi-asserted-by":"publisher","DOI":"10.1109\/MM.2018.112130359"},{"key":"ref185","article-title":"Energy-efficient optical neural network enabled by weight reuse and blend","author":"Li","year":"2024","journal-title":"arXiv:2409.01836"},{"key":"ref186","doi-asserted-by":"publisher","DOI":"10.1126\/science.adl1203"},{"key":"ref187","doi-asserted-by":"publisher","DOI":"10.1038\/nature23474"},{"key":"ref188","doi-asserted-by":"publisher","DOI":"10.1109\/ISCA.2016.12"},{"key":"ref189","doi-asserted-by":"publisher","DOI":"10.1109\/TSUSC.2022.3190242"},{"key":"ref190","doi-asserted-by":"publisher","DOI":"10.1007\/s00521-023-08345-y"},{"key":"ref191","doi-asserted-by":"publisher","DOI":"10.1109\/TC.2024.3449108"},{"key":"ref192","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR42600.2020.00215"},{"key":"ref193","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v39i20.35415"},{"key":"ref194","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-01249-6_18"},{"key":"ref195","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00293"},{"key":"ref196","article-title":"ProxylessNAS: Direct neural architecture search on target task and hardware","volume-title":"Proc. Int. Conf. Learn. Represent. (ICLR)","author":"Cai"},{"key":"ref197","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.01099"},{"key":"ref198","article-title":"Hw-NAS-bench: Hardware-aware neural architecture search benchmark","volume-title":"Proc. Int. Conf. Learn. Represent. (ICLR)","author":"Li"},{"key":"ref199","first-page":"3505","article-title":"Auto-NBA: Efficient and effective search over the joint space of networks, bitwidths, and accelerators","volume-title":"Proc. 38th Int. Conf. Mach. Learn. (ICML)","author":"Fu"},{"key":"ref200","doi-asserted-by":"publisher","DOI":"10.1145\/3575798"},{"key":"ref201","doi-asserted-by":"publisher","DOI":"10.1109\/ACCESS.2023.3320133"},{"key":"ref202","article-title":"MCUNet: Tiny deep learning on IoT devices","volume-title":"Proc. Adv. Neural Inf. Process. Syst., Annu. Conf. Neural Inf. Process. Syst.","author":"Lin"},{"key":"ref203","article-title":"CMSIS-NN: Efficient neural network kernels for arm Cortex-M CPUs","author":"Lai","year":"2018","journal-title":"arXiv:1801.06601"},{"key":"ref204","article-title":"TensorFlow lite micro: Embedded machine learning on TinyML systems","author":"David","year":"2020","journal-title":"arXiv:2010.08678"},{"key":"ref205","doi-asserted-by":"publisher","DOI":"10.1145\/3369837"},{"key":"ref206","doi-asserted-by":"publisher","DOI":"10.1145\/1950365.1950386"},{"key":"ref207","doi-asserted-by":"publisher","DOI":"10.1145\/3133920"},{"key":"ref208","doi-asserted-by":"publisher","DOI":"10.1109\/dasc\/picom\/cbdcom\/cy59711.2023.10361321"},{"key":"ref209","doi-asserted-by":"publisher","DOI":"10.1109\/icc.2019.8761315"},{"key":"ref210","first-page":"19","article-title":"Oort: Efficient federated learning via guided participant selection","volume-title":"Proc. 15th USENIX Symp. Operating Syst. Design Implement. (OSDI)","author":"Lai"},{"key":"ref211","doi-asserted-by":"publisher","DOI":"10.1109\/TMC.2019.2947893"},{"key":"ref212","doi-asserted-by":"publisher","DOI":"10.1145\/3600006.3613165"},{"key":"ref213","doi-asserted-by":"publisher","DOI":"10.52202\/068431-1189"},{"key":"ref214","doi-asserted-by":"publisher","DOI":"10.1109\/sc41405.2020.00024"},{"key":"ref215","article-title":"Megatron-LM: Training multi-billion parameter language models using model parallelism","author":"Shoeybi","year":"2019","journal-title":"arXiv:1909.08053"},{"key":"ref216","doi-asserted-by":"publisher","DOI":"10.1109\/ACCESS.2021.3083697"},{"key":"ref217","article-title":"MLIR: A compiler infrastructure for the end of Moore\u2019s Law","author":"Lattner","year":"2020","journal-title":"arXiv:2002.11054"},{"key":"ref218","doi-asserted-by":"publisher","DOI":"10.1126\/science.adh1174"},{"key":"ref219","article-title":"GREEN: Carbon-efficient resource scheduling for machine learning clusters","volume-title":"Proc. 22nd USENIX Symp. Networked Syst. Design Implement. (NSDI)","author":"Xu"}],"container-title":["IEEE Access"],"original-title":[],"link":[{"URL":"https:\/\/ieeexplore.ieee.org\/ielam\/6287639\/11323511\/11692858-aam.pdf","content-type":"application\/pdf","content-version":"am","intended-application":"syndication"},{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/6287639\/11323511\/11692858.pdf?arnumber=11692858","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,24]],"date-time":"2026-09-24T19:08:07Z","timestamp":1790276887000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11692858\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026]]},"references-count":219,"URL":"https:\/\/doi.org\/10.1109\/access.2026.3734236","relation":{},"ISSN":["2169-3536"],"issn-type":[{"value":"2169-3536","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026]]}}}