{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,4,25]],"date-time":"2026-04-25T08:36:44Z","timestamp":1777106204125,"version":"3.51.4"},"reference-count":45,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","issue":"4","license":[{"start":{"date-parts":[[2025,4,1]],"date-time":"2025-04-01T00:00:00Z","timestamp":1743465600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2025,4,1]],"date-time":"2025-04-01T00:00:00Z","timestamp":1743465600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2025,4,1]],"date-time":"2025-04-01T00:00:00Z","timestamp":1743465600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"DOI":"10.13039\/501100012166","name":"National Key Research and Development Program of China","doi-asserted-by":"publisher","award":["2022YFB4501000"],"award-info":[{"award-number":["2022YFB4501000"]}],"id":[{"id":"10.13039\/501100012166","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","award":["62202328"],"award-info":[{"award-number":["62202328"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE Trans. Comput."],"published-print":{"date-parts":[[2025,4]]},"DOI":"10.1109\/tc.2025.3528125","type":"journal-article","created":{"date-parts":[[2025,1,10]],"date-time":"2025-01-10T20:17:51Z","timestamp":1736540271000},"page":"1431-1445","source":"Crossref","is-referenced-by-count":2,"title":["SLOpt: Serving Real-Time Inference Pipeline With Strict Latency Constraint"],"prefix":"10.1109","volume":"74","author":[{"ORCID":"https:\/\/orcid.org\/0009-0009-7865-4905","authenticated-orcid":false,"given":"Zhixin","family":"Zhao","sequence":"first","affiliation":[{"name":"Tianjin Key Laboratory of Advanced Networking, Department of Intelligence and Computing, Tianjin University, Tianjin, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0004-0458-0900","authenticated-orcid":false,"given":"Yitao","family":"Hu","sequence":"additional","affiliation":[{"name":"Tianjin Key Laboratory of Advanced Networking, Department of Intelligence and Computing, Tianjin University, Tianjin, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0003-6392-6468","authenticated-orcid":false,"given":"Guotao","family":"Yang","sequence":"additional","affiliation":[{"name":"Tianjin Key Laboratory of Advanced Networking, Department of Intelligence and Computing, Tianjin University, Tianjin, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0002-1699-8592","authenticated-orcid":false,"given":"Ziqi","family":"Gong","sequence":"additional","affiliation":[{"name":"Tianjin Key Laboratory of Advanced Networking, Department of Intelligence and Computing, Tianjin University, Tianjin, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0002-1444-1272","authenticated-orcid":false,"given":"Chen","family":"Shen","sequence":"additional","affiliation":[{"name":"Tianjin Key Laboratory of Advanced Networking, Department of Intelligence and Computing, Tianjin University, Tianjin, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-1967-2192","authenticated-orcid":false,"given":"Laiping","family":"Zhao","sequence":"additional","affiliation":[{"name":"Tianjin Key Laboratory of Advanced Networking, Department of Intelligence and Computing, Tianjin University, Tianjin, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-8507-0339","authenticated-orcid":false,"given":"Wenxin","family":"Li","sequence":"additional","affiliation":[{"name":"Tianjin Key Laboratory of Advanced Networking, Department of Intelligence and Computing, Tianjin University, Tianjin, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4746-5599","authenticated-orcid":false,"given":"Xiulong","family":"Liu","sequence":"additional","affiliation":[{"name":"Tianjin Key Laboratory of Advanced Networking, Department of Intelligence and Computing, Tianjin University, Tianjin, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-4817-5187","authenticated-orcid":false,"given":"Wenyu","family":"Qu","sequence":"additional","affiliation":[{"name":"Tianjin Key Laboratory of Advanced Networking, Department of Intelligence and Computing, Tianjin University, Tianjin, China"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"263","reference":[{"key":"ref1","article-title":"The shift from models to compound AI systems","author":"Zaharia","year":"2024"},{"key":"ref2","article-title":"Does one large model rule them all? Predictions on the future AI ecosystem","author":"Raghu","year":"2023"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.1145\/3472883.3486993"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.1145\/3341301.3359658"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1145\/3419111.3421285"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.1145\/3623278.3624753"},{"key":"ref7","article-title":"Live video analytics at scale with approximation and delay-tolerance","author":"Zhang","year":"2017","journal-title":"14th USENIX Symp. Networked Syst. Des. Implementation"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1145\/2408776.2408794"},{"key":"ref9","article-title":"$\\{${Model-Switching$\\}$}: Dealing with fluctuating workloads in $\\{${Machine-learning-as-a-service$\\}$} systems","volume-title":"Proc. 12th USENIX Workshop Hot Topics Cloud Comput. (HotCloud 20)","author":"Zhang","year":"2020"},{"key":"ref10","first-page":"1049","article-title":"MArk: Exploiting cloud services for cost-effective, slo-aware machine learning inference serving","author":"Zhang","year":"2019","journal-title":"USENIX Annu. Tech. Conf."},{"key":"ref11","first-page":"1041","article-title":"Cocktail: A multidimensional optimization for model serving in cloud","volume-title":"Proc. USENIX NSDI","author":"Gunasekaran","year":"2022"},{"key":"ref12","first-page":"443","article-title":"Serving DNNs like clockwork: Performance predictability from the bottom up","volume-title":"Proc. 14th USENIX Symp.Oper. Syst. Design Implement. (OSDI)","author":"Gujarati","year":"2020"},{"key":"ref13","first-page":"787","article-title":"SHEPHERD: Serving DNNs in the wild","volume-title":"Proc. 20th USENIX Symposium on Networked Systems Design and Implementation (NSDI)","author":"Zhang","year":"2023"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.1109\/GHTC.2018.8601887"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.1016\/j.comnet.2009.02.019"},{"key":"ref16","first-page":"613","article-title":"Clipper: A low-latency online prediction serving system.\u201d","volume-title":"Proc. NSDI","volume":"17","author":"Crankshaw","year":"2017"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1109\/ACCESS.2021.3097964"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1145\/3503222.3507750"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.1145\/3503222.3507709"},{"key":"ref20","article-title":"Tensorflow-serving: Flexible, high-performance ml serving","author":"Olston","year":"2017"},{"issue":"2016","key":"ref21","first-page":"265","article-title":"TensorFlow: A system for large-scale machine learning","volume-title":"OSDI","volume":"16","author":"Abadi","year":"2016"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.5555\/3454287.3455008"},{"key":"ref23","doi-asserted-by":"publisher","DOI":"10.1109\/DAC18072.2020.9218700"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.1016\/j.ijforecast.2019.07.001"},{"key":"ref25","first-page":"1","article-title":"Pollux: Co-adaptive cluster scheduling for goodput-optimized deep learning.\u201d","volume":"21","author":"Qiao","year":"2021","journal-title":"OSDI"},{"key":"ref26","article-title":"Amazon Web Services: Set the default instance warmup","year":"2024"},{"key":"ref27","article-title":"Amazon EC2 Spot Instances","year":"2024"},{"key":"ref28","article-title":"Amazon Web Services","year":"2024"},{"key":"ref29","article-title":"Linux Virtual Machines Pricing","year":"2024"},{"key":"ref30","article-title":"Spot Virtual Machines","year":"2024"},{"key":"ref31","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2018.00907"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-46448-0_2"},{"key":"ref33","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2016.90"},{"key":"ref34","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2018.00474"},{"key":"ref35","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v31i1.11231"},{"key":"ref36","doi-asserted-by":"publisher","DOI":"10.1109\/TCC.2020.3006751"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.1145\/3642970.3655833"},{"key":"ref38","doi-asserted-by":"publisher","DOI":"10.1007\/s11241-024-09418-4"},{"key":"ref39","doi-asserted-by":"publisher","DOI":"10.1007\/s11036-023-02131-9"},{"key":"ref40","doi-asserted-by":"publisher","DOI":"10.1145\/3460352"},{"key":"ref41","doi-asserted-by":"publisher","DOI":"10.1145\/3423211.3425690"},{"key":"ref42","first-page":"199","article-title":"Serving heterogeneous machine learning models on multi-GPU servers with spatio-temporal sharing","volume-title":"Proc. USENIX Annu. Tech. Conf. (USENIX ATC)","author":"Choi","year":"2022"},{"key":"ref43","article-title":"Nvidia tensorrt, an sdk for high-performance deep learning inference","year":"2023"},{"key":"ref44","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2018.00286"},{"key":"ref45","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2017.298"}],"container-title":["IEEE Transactions on Computers"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/12\/10924434\/10836842.pdf?arnumber=10836842","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,4,15]],"date-time":"2025-04-15T17:35:51Z","timestamp":1744738551000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/10836842\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,4]]},"references-count":45,"journal-issue":{"issue":"4"},"URL":"https:\/\/doi.org\/10.1109\/tc.2025.3528125","relation":{},"ISSN":["0018-9340","1557-9956","2326-3814"],"issn-type":[{"value":"0018-9340","type":"print"},{"value":"1557-9956","type":"electronic"},{"value":"2326-3814","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,4]]}}}