{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,26]],"date-time":"2026-08-26T15:39:39Z","timestamp":1787758779547,"version":"build-2784847793"},"reference-count":61,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","issue":"3","license":[{"start":{"date-parts":[[2025,6,1]],"date-time":"2025-06-01T00:00:00Z","timestamp":1748736000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2025,6,1]],"date-time":"2025-06-01T00:00:00Z","timestamp":1748736000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2025,6,1]],"date-time":"2025-06-01T00:00:00Z","timestamp":1748736000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"DOI":"10.13039\/501100001809","name":"NSFC Project","doi-asserted-by":"publisher","award":["62371432"],"award-info":[{"award-number":["62371432"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100001809","name":"NSFC Project","doi-asserted-by":"publisher","award":["62201627"],"award-info":[{"award-number":["62201627"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE Trans. Netw."],"published-print":{"date-parts":[[2025,6]]},"DOI":"10.1109\/ton.2025.3526228","type":"journal-article","created":{"date-parts":[[2025,1,10]],"date-time":"2025-01-10T15:17:34Z","timestamp":1736522254000},"page":"1236-1251","source":"Crossref","is-referenced-by-count":20,"title":["P4INC-AOI: All-Optical Interconnect Empowered by In-Network Computing for DML Workloads"],"prefix":"10.1109","volume":"33","author":[{"ORCID":"https:\/\/orcid.org\/0009-0003-4292-1123","authenticated-orcid":false,"given":"Xuexia","family":"Xie","sequence":"first","affiliation":[{"name":"School of Information Science and Technology, University of Science and Technology of China, Hefei, Anhui, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0001-3956-3244","authenticated-orcid":false,"given":"Binjun","family":"Tang","sequence":"additional","affiliation":[{"name":"School of Information Science and Technology, University of Science and Technology of China, Hefei, Anhui, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-7805-6237","authenticated-orcid":false,"given":"Xiaoliang","family":"Chen","sequence":"additional","affiliation":[{"name":"School of Information Science and Technology, University of Science and Technology of China, Hefei, Anhui, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4251-788X","authenticated-orcid":false,"given":"Zuqing","family":"Zhu","sequence":"additional","affiliation":[{"name":"School of Information Science and Technology, University of Science and Technology of China, Hefei, Anhui, China"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.1016\/j.infsof.2023.107268"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.1109\/TNSM.2017.2711610"},{"key":"ref3","volume-title":"OpenAI GPT-3: Everything You Need To Know [Updated]","author":"Cooper","year":"2023"},{"key":"ref4","first-page":"1223","article-title":"Large scale distributed deep networks","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Dean"},{"key":"ref5","first-page":"265","article-title":"TensorFlow: A system for large-scale machine learning","volume-title":"Proc. 12th USENIX Symp. Oper. Syst. Design Implement.","author":"Abadi"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.5555\/3291168.3291210"},{"key":"ref7","article-title":"Communication-efficient distributed deep learning: A comprehensive survey","author":"Tang","year":"2020","journal-title":"arXiv:2003.06307"},{"key":"ref8","first-page":"181","article-title":"Poseidon: An efficient communication architecture for distributed deep learning on GPU clusters","volume-title":"Proc. USENIX ATC","author":"Zhang"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1145\/3267809.3267840"},{"key":"ref10","first-page":"463","article-title":"A unified architecture for accelerating distributed DNN training in heterogeneous GPU\/CPU clusters","volume-title":"Proc. 14th USENIX Symp. Operating Syst. Design Implement. (OSDI)","author":"Jiang"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1145\/1402958.1402967"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.1109\/MNET.2015.7293303"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1109\/JLT.2017.2777605"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.1109\/MCOM.001.1800157"},{"key":"ref15","article-title":"Gemini: Practical reconfigurable datacenter networks with topology and traffic engineering","author":"Zhang","year":"2021","journal-title":"arXiv:2110.08374"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.1145\/1592568.1592576"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1109\/ACCESS.2019.2924332"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1145\/3579371.3589350"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.1145\/1851182.1851223"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.1109\/TNET.2013.2253120"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.1109\/JLT.2012.2227683"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.1364\/JOCN.5.000836"},{"key":"ref23","doi-asserted-by":"publisher","DOI":"10.1364\/JOCN.5.00A100"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.1109\/JLT.2015.2479366"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.1109\/JLT.2013.2294389"},{"key":"ref26","first-page":"1","article-title":"What you need to know about optical circuit reconfigurations in datacenter networks","volume-title":"Proc. ITC","author":"Zerwas"},{"key":"ref27","doi-asserted-by":"publisher","DOI":"10.1109\/MEMS58180.2024.10439411"},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.1145\/2486001.2486007"},{"key":"ref29","doi-asserted-by":"publisher","DOI":"10.1145\/3387514.3406221"},{"key":"ref30","doi-asserted-by":"publisher","DOI":"10.1109\/JSTQE.2019.2950770"},{"key":"ref31","doi-asserted-by":"publisher","DOI":"10.1145\/3544216.3544265"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.1145\/3491054"},{"key":"ref33","doi-asserted-by":"publisher","DOI":"10.1145\/2656877.2656890"},{"key":"ref34","volume-title":"In-Network Computing","author":"Zilberman","year":"2019"},{"key":"ref35","doi-asserted-by":"publisher","DOI":"10.1364\/OFC.2023.Th1D.4"},{"key":"ref36","doi-asserted-by":"publisher","DOI":"10.1016\/j.knosys.2020.106002"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.1145\/3386367.3432588"},{"key":"ref38","first-page":"785","article-title":"Scaling distributed machine learning with in-network aggregation","volume-title":"Proc. 18th USENIX Symp. Netw. Syst. Design Implement. (NSDI)","author":"Sapio"},{"key":"ref39","first-page":"741","article-title":"ATP: In-network aggregation for multi-tenant learning","volume-title":"Proc. 18th USENIX Symp. Netw. Syst. Design Implement. (NSDI)","author":"Lao"},{"key":"ref40","first-page":"2595","article-title":"Parallelized stochastic gradient descent","volume-title":"Proc. NeurIPS","volume":"23","author":"Zinkevich"},{"key":"ref41","doi-asserted-by":"publisher","DOI":"10.1017\/cbo9781107298019"},{"key":"ref42","doi-asserted-by":"publisher","DOI":"10.1145\/3377454"},{"key":"ref43","doi-asserted-by":"publisher","DOI":"10.1016\/j.jpdc.2008.09.002"},{"key":"ref44","doi-asserted-by":"publisher","DOI":"10.1145\/3377454"},{"key":"ref45","doi-asserted-by":"publisher","DOI":"10.1109\/ICNP61940.2024.10858570"},{"key":"ref46","doi-asserted-by":"publisher","DOI":"10.1109\/TNET.2022.3210534"},{"key":"ref47","doi-asserted-by":"publisher","DOI":"10.1109\/TNET.2023.3265276"},{"key":"ref48","article-title":"Domino: Eliminating communication in LLM training via generic tensor slicing and overlapping","author":"Wang","year":"2024","journal-title":"arXiv:2409.15241"},{"key":"ref49","volume-title":"NetFPGA","year":"2024"},{"key":"ref50","doi-asserted-by":"publisher","DOI":"10.1145\/3452296.3472940"},{"key":"ref51","volume-title":"Intel Tofino","year":"2024"},{"key":"ref52","doi-asserted-by":"publisher","DOI":"10.1145\/3485983.3494853"},{"key":"ref53","first-page":"829","article-title":"In-network aggregation for shared machine learning clusters","volume-title":"Proc. MLSys","volume":"3","author":"Gebara"},{"key":"ref54","doi-asserted-by":"publisher","DOI":"10.1109\/TNET.2023.3244794"},{"key":"ref55","volume-title":"Intel Intelligent Fabric Processors","year":"2024"},{"key":"ref56","doi-asserted-by":"publisher","DOI":"10.1109\/ACCESS.2021.3086704"},{"key":"ref57","first-page":"945","article-title":"MLaaS in the wild: Workload analysis and scheduling in large-scale heterogeneous GPU clusters","volume-title":"Proc. NSDI","author":"Weng"},{"key":"ref58","volume-title":"Computers and Intractability: A Guide to the Theory of NP-Completeness","author":"Garey","year":"1979"},{"key":"ref59","doi-asserted-by":"publisher","DOI":"10.1145\/2999572.2999592"},{"key":"ref60","first-page":"947","article-title":"Analysis of large-scale multi-tenant GPU clusters for DNN training workloads","volume-title":"Proc. USENIX ATC","author":"Jeon"},{"key":"ref61","volume-title":"Coflow-Benchmark","year":"2024"}],"container-title":["IEEE Transactions on Networking"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/10723154\/11039001\/10836894.pdf?arnumber=10836894","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,6,18]],"date-time":"2025-06-18T13:37:32Z","timestamp":1750253852000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/10836894\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,6]]},"references-count":61,"journal-issue":{"issue":"3"},"URL":"https:\/\/doi.org\/10.1109\/ton.2025.3526228","relation":{},"ISSN":["2998-4157"],"issn-type":[{"value":"2998-4157","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,6]]}}}