{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,5]],"date-time":"2026-08-05T06:46:18Z","timestamp":1785912378194,"version":"3.56.0"},"reference-count":50,"publisher":"IEEE","license":[{"start":{"date-parts":[[2026,6,1]],"date-time":"2026-06-01T00:00:00Z","timestamp":1780272000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2026,6,1]],"date-time":"2026-06-01T00:00:00Z","timestamp":1780272000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2026,6,1]],"date-time":"2026-06-01T00:00:00Z","timestamp":1780272000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2026,6]]},"DOI":"10.1109\/2575-8411.2026.00048","type":"proceedings-article","created":{"date-parts":[[2026,8,4]],"date-time":"2026-08-04T19:08:43Z","timestamp":1785870523000},"page":"436-446","source":"Crossref","is-referenced-by-count":0,"title":["HiAsCC: Hierarchical Asynchronous Collective Communication Method for Large Model Training"],"prefix":"10.1109","author":[{"given":"Zhihang","family":"Tang","sequence":"first","affiliation":[{"name":"College of Intelligence and Computing, Tianjin University"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Chenxi","family":"Li","sequence":"additional","affiliation":[{"name":"School of Computer Science, Beijing University of Posts and Telecommunications"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Bo","family":"He","sequence":"additional","affiliation":[{"name":"School of Computer Science, Beijing University of Posts and Telecommunications"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Qi","family":"Qi","sequence":"additional","affiliation":[{"name":"School of Computer Science, Beijing University of Posts and Telecommunications"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yulong","family":"Tao","sequence":"additional","affiliation":[{"name":"School of Computer Science, Beijing University of Posts and Telecommunications"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jingyu","family":"Wang","sequence":"additional","affiliation":[{"name":"School of Computer Science, Beijing University of Posts and Telecommunications"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Laiping","family":"Zhao","sequence":"additional","affiliation":[{"name":"College of Intelligence and Computing, Tianjin University"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Keqiu","family":"Li","sequence":"additional","affiliation":[{"name":"College of Intelligence and Computing, Tianjin University"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.1145\/3744746"},{"key":"ref2","article-title":"Qwen3 technical report","author":"Yang","year":"2025","journal-title":"arXiv preprint"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.1145\/3731569.3764848"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.1145\/3760789"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1145\/3567955.3567959"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.1109\/MM.2021.3091475"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1109\/IPDPS64566.2025.00089"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1145\/3723851.3723853"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1145\/3731599.3767453"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.1016\/j.jpdc.2008.09.002"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1109\/OJCOMS.2023.3349372"},{"key":"ref12","first-page":"667","article-title":"Autoccl: Automated collective communication tuning for accelerating distributed and parallel dnn training","volume-title":"22nd USENIX Symposium on Networked Systems Design and Implementation (NSDI 25)","author":"Xu","year":"2025"},{"key":"ref13","article-title":"Understanding intra-node communication in hpc systems and datacenters","author":"Tarraga-Moreno","year":"2025","journal-title":"arXiv preprint"},{"key":"ref14","first-page":"739","article-title":"Topoopt: Co-optimizing network topology and parallelization strategy for distributed training jobs","volume-title":"20th USENIX Symposium on Networked Systems Design and Implementation (NSDI 23)","author":"Wang","year":"2023"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.1109\/MICRO61859.2024.00068"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.1145\/3696410.3714741"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1109\/IWQoS65803.2025.11143446"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1145\/3470496.3527382"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.1145\/3620666.3651379"},{"key":"ref20","article-title":"Co2: Efficient distributed training with full communicationcomputation overlap","author":"Sun","year":"2024","journal-title":"arXiv preprint"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.1145\/3650200.3656636"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.1109\/ICDCS57875.2023.00015"},{"key":"ref23","doi-asserted-by":"publisher","DOI":"10.1109\/ICDCS63083.2025.00094"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.52202\/079017-0884"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.5555\/2685048.2685095"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1109\/TNNLS.2021.3084806"},{"key":"ref27","doi-asserted-by":"publisher","DOI":"10.1109\/CCGrid59990.2024.00043"},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.1109\/ACCESS.2023.3347928"},{"key":"ref29","first-page":"344","article-title":"Stochastic gradient push for distributed deep learning","volume-title":"Proceedings of the 36th International Conference on Machine Learning","author":"Assran","year":"2019"},{"key":"ref30","doi-asserted-by":"publisher","DOI":"10.1109\/VTC2025-Spring65109.2025.11174740"},{"key":"ref31","first-page":"705","article-title":"Efficient {Direct-Connect} topologies for collective communications","volume-title":"22nd USENIX Symposium on Networked Systems Design and Implementation (NSDI 25)","author":"Zhao","year":"2025"},{"key":"ref32","article-title":"Blueconnect: Novel hierarchical allreduce on multi-tired network for deep learning","volume-title":"Proceedings of the 2nd SysML Conference","author":"Cho","year":"2019"},{"key":"ref33","doi-asserted-by":"publisher","DOI":"10.1109\/TON.2025.3556013"},{"key":"ref34","doi-asserted-by":"publisher","DOI":"10.1109\/HPCA61900.2025.00114"},{"key":"ref35","article-title":"Optimizing allreduce operations for heterogeneous architectures with multiple processes per gpu","author":"Adams","year":"2025","journal-title":"arXiv preprint"},{"key":"ref36","doi-asserted-by":"publisher","DOI":"10.1109\/ICDCS60910.2024.00012"},{"key":"ref37","first-page":"463","article-title":"A unified architecture for accelerating distributed DNN training in heterogeneous GPU\/CPU clusters","volume-title":"14th USENIX Symposium on Operating Systems Design and Implementation (OSDI 20)","author":"Jiang","year":"2020"},{"key":"ref38","doi-asserted-by":"publisher","DOI":"10.14778\/3561261.3561265"},{"key":"ref39","article-title":"Highly scalable deep learning training system with mixed-precision: Training imagenet in four minutes","author":"Jia","year":"2018","journal-title":"arXiv preprint"},{"key":"ref40","article-title":"Exponential moving average of weights in deep learning: Dynamics and benefits","author":"Morales-Brotons","year":"2024","journal-title":"arXiv preprint"},{"key":"ref41","doi-asserted-by":"publisher","DOI":"10.1007\/s10107-022-01822-7"},{"key":"ref42","doi-asserted-by":"publisher","DOI":"10.1145\/3394486.3406703"},{"key":"ref43","doi-asserted-by":"publisher","DOI":"10.1109\/SC41405.2020.00024"},{"key":"ref44","author":"Zhang","year":"2022","journal-title":"Opt: Open pre-trained transformer language models"},{"key":"ref45","article-title":"Llama3.1-foundationai-securityllm-base-8b technical report","author":"Kassianik","year":"2025","journal-title":"arXiv preprint"},{"key":"ref46","volume-title":"rm-static dataset","year":"2023"},{"key":"ref47","volume-title":"full-hh-rlhf dataset","year":"2023"},{"key":"ref48","volume-title":"synthetic-instruct-gptj-pairwise dataset","year":"2023"},{"key":"ref49","volume-title":"rlhf-reward-datasets dataset","year":"2023"},{"key":"ref50","article-title":"Deep reinforcement learning from human preferences","author":"Christiano","year":"2017","journal-title":"Advances in Neural Information Processing Systems"}],"event":{"name":"2026 IEEE 46th International Conference on Distributed Computing Systems (ICDCS)","location":"Seoul, Korea, Republic of","start":{"date-parts":[[2026,6,22]]},"end":{"date-parts":[[2026,6,25]]}},"container-title":["2026 IEEE 46th International Conference on Distributed Computing Systems (ICDCS)"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/11618978\/11618981\/11619228.pdf?arnumber=11619228","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,5]],"date-time":"2026-08-05T05:57:47Z","timestamp":1785909467000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11619228\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,6]]},"references-count":50,"URL":"https:\/\/doi.org\/10.1109\/2575-8411.2026.00048","relation":{},"subject":[],"published":{"date-parts":[[2026,6]]}}}