{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,3,5]],"date-time":"2026-03-05T07:14:24Z","timestamp":1772694864006,"version":"3.50.1"},"reference-count":65,"publisher":"IEEE","license":[{"start":{"date-parts":[[2026,1,31]],"date-time":"2026-01-31T00:00:00Z","timestamp":1769817600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2026,1,31]],"date-time":"2026-01-31T00:00:00Z","timestamp":1769817600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"DOI":"10.13039\/501100001667","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","award":["U22A6001"],"award-info":[{"award-number":["U22A6001"]}],"id":[{"id":"10.13039\/501100001667","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2026,1,31]]},"DOI":"10.1109\/hpca68181.2026.11408533","type":"proceedings-article","created":{"date-parts":[[2026,3,4]],"date-time":"2026-03-04T20:47:22Z","timestamp":1772657242000},"page":"1-17","source":"Crossref","is-referenced-by-count":0,"title":["AutoHAAP: Automated Heterogeneity-Aware Asymmetric Partitioning for LLM Training"],"prefix":"10.1109","author":[{"given":"Yuanyuan","family":"Wang","sequence":"first","affiliation":[{"name":"Zhejiang Lab"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Nana","family":"Tang","sequence":"additional","affiliation":[{"name":"Zhejiang Lab"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Yuyang","family":"Wang","sequence":"additional","affiliation":[{"name":"Zhejiang Lab"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Shu","family":"Pan","sequence":"additional","affiliation":[{"name":"Zhejiang Lab"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Dingding","family":"Yu","sequence":"additional","affiliation":[{"name":"Zhejiang Lab"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Zeyue","family":"Wang","sequence":"additional","affiliation":[{"name":"Zhejiang Lab"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Mou","family":"Sun","sequence":"additional","affiliation":[{"name":"Zhejiang Lab"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Kejie","family":"Fu","sequence":"additional","affiliation":[{"name":"Zhejiang Lab"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Fangyu","family":"Wang","sequence":"additional","affiliation":[{"name":"Zhejiang Lab"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Yunchuan","family":"Chen","sequence":"additional","affiliation":[{"name":"Zhejiang Lab"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Ning","family":"Sun","sequence":"additional","affiliation":[{"name":"Zhejiang Lab"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Fei","family":"Yang","sequence":"additional","affiliation":[{"name":"Zhejiang Lab"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.1038\/s41586-024-07487-w"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.1038\/s41586-023-06185-3"},{"key":"ref3","article-title":"Language models are few-shot learners","volume-title":"in Proceedings of the 34th International Conference on Neural Information Processing Systems, ser. NIPS \u201920. Red Hook","author":"Brown"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.1109\/TKDE.2025.3554028"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1145\/3502181.3531462"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.1145\/3695053.3731410"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1038\/s43856-023-00370-1"},{"key":"ref8","article-title":"Deepseek-v3 technical report","volume-title":"arXiv preprint","author":"Liu","year":"2024"},{"key":"ref9","doi-asserted-by":"crossref","first-page":"12619","DOI":"10.18653\/v1\/2025.findings-acl.653","article-title":"Maximum score routing for mixture-of-experts","volume-title":"in Findings of the Association for Computational Linguistics: ACL 2025","author":"Dong","year":"2025"},{"key":"ref10","article-title":"Efficient training of large language models on distributed infrastructures: A survey","author":"Duan","year":"2024","journal-title":"arXiv preprint"},{"key":"ref11","volume-title":"The llama 3 herd of models","author":"Dubey","year":"2024"},{"key":"ref12","volume-title":"Fairscale: A general purpose modular pytorch library for high performance and large scale training","year":"2021"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1145\/3721145.3730418"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.1145\/3600006.3613175"},{"key":"ref15","first-page":"673","article-title":"Whale: Efficient giant model training over heterogeneous GPUs","volume-title":"in 2022 USENIX Annual Technical Conference (USENIX ATC 22)","author":"Jia"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.1145\/3035918.3035933"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1109\/CCGRID.2019.00053"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1038\/s41586-021-03819-2"},{"key":"ref19","article-title":"Scaling laws for neural language models","author":"Kaplan","year":"2020","journal-title":"arXiv preprint"},{"key":"ref20","volume-title":"Ai 202","author":"Kokotajlo","year":"2025"},{"key":"ref21","first-page":"341","article-title":"Reducing activation recomputation in large transformer models","volume-title":"in Proceedings of Machine Learning and Systems","volume":"5","author":"Korthikanti"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.1007\/s10462-024-10888-y"},{"key":"ref23","doi-asserted-by":"publisher","DOI":"10.1109\/TPDS.2023.3247001"},{"key":"ref24","first-page":"6630","article-title":"Amp: Automatically finding model parallel strategies with heterogeneity awareness","volume-title":"in Advances in Neural Information Processing Systems","volume":"35","author":"Li"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.1145\/3605573.3605613"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1145\/3458817.3476145"},{"key":"ref27","article-title":"Understanding stragglers in large model training using what-if analysis","volume-title":"in Proceedings of the 19th USENIX Conference on Operating Systems Design and Implementation, ser. OSDI \u201925. USA: USENIX Association","author":"Lin"},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.1109\/HPCA56546.2023.10071043"},{"key":"ref29","doi-asserted-by":"publisher","DOI":"10.1109\/HPCA57654.2024.00067"},{"key":"ref30","doi-asserted-by":"publisher","DOI":"10.1145\/3627703.3629554"},{"key":"ref31","doi-asserted-by":"publisher","DOI":"10.1145\/3297858.3304009"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.14778\/3570690.3570697"},{"key":"ref33","doi-asserted-by":"publisher","DOI":"10.3390\/info16080688"},{"key":"ref34","doi-asserted-by":"publisher","DOI":"10.1145\/3620665.3640375"},{"key":"ref35","doi-asserted-by":"publisher","DOI":"10.1145\/3341301.3359646"},{"key":"ref36","first-page":"79377947","article-title":"Memory-efficient pipeline-parallel dnn training","volume-title":"in International Conference on Machine Learning (ICML2021)","author":"Narayanan"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.1145\/3458817.3476209"},{"key":"ref38","article-title":"GPT-4 technical report","volume-title":"CoRR, vol. abs\/2303.08774","year":"2023"},{"key":"ref39","article-title":"Zero bubble (almost) pipeline parallelism","volume-title":"in The Twelfth International Conference on Learning Representations","author":"Qi"},{"key":"ref40","doi-asserted-by":"publisher","DOI":"10.1145\/3394486.3406703"},{"key":"ref41","volume-title":"Megatron-lm: Training multi-billion parameter language models using model parallelism","author":"Shoeybi","year":"2020"},{"key":"ref42","doi-asserted-by":"publisher","DOI":"10.1038\/s41591-024-03423-7"},{"key":"ref43","article-title":"CO2: Efficient distributed training with full communication-computation overlap","volume-title":"in The Twelfth International Conference on Learning Representations","author":"Sun"},{"key":"ref44","doi-asserted-by":"publisher","DOI":"10.1145\/3620666.3651359"},{"key":"ref45","article-title":"Pangu pro moe: Mixture of grouped experts for efficient sparsity","author":"Tang","year":"2025","journal-title":"arXiv preprint"},{"key":"ref46","first-page":"563","article-title":"Metis: Fast automatic distributed training on heterogeneous GPUs","volume-title":"in 2024 USENIX Annual Technical Conference (USENIX ATC 24)","author":"Um"},{"key":"ref47","doi-asserted-by":"publisher","DOI":"10.1145\/3620666.3651357"},{"key":"ref48","article-title":"Astra: Efficient and money-saving automatic parallel strategies search on heterogeneous gpus","author":"Wang","year":"2025","journal-title":"arXiv preprint"},{"key":"ref49","doi-asserted-by":"publisher","DOI":"10.1145\/3567955.3567959"},{"key":"ref50","doi-asserted-by":"publisher","DOI":"10.1109\/tkde.2024.3370614"},{"key":"ref51","doi-asserted-by":"publisher","DOI":"10.1145\/3676641.3715998"},{"key":"ref52","doi-asserted-by":"publisher","DOI":"10.1109\/tpami.2021.3107796"},{"key":"ref53","first-page":"arXiv","article-title":"Hethub: A heterogeneous distributed hybrid training system for large-scale models","author":"Xu","year":"2024","journal-title":"arXiv e-prints"},{"key":"ref54","article-title":"Flashflex: Accommodating large language model training over heterogeneous environment","volume-title":"ArXiv, vol. abs\/2409.01143","author":"Yan","year":"2024"},{"key":"ref55","article-title":"Qwen2.5 technical report","volume-title":"CoRR, vol. abs\/2412.15115","author":"Yang","year":"2024"},{"key":"ref56","doi-asserted-by":"publisher","DOI":"10.1145\/3673038.3673095"},{"key":"ref57","doi-asserted-by":"crossref","DOI":"10.23919\/DATE58400.2024.10546826","article-title":"Pipette: Automatic fine-grained large language model training configurator for real-world clusters","volume-title":"Institute of Electrical and Electronics Engineers Inc., 2024, Conference paper, cited by: 0; Conference name: 2024 Design, Automation and Test in Europe Conference and Exhibition, DATE 2024","author":"Yim"},{"key":"ref58","first-page":"545","article-title":"Accelerating the training of large language models using efficient activation rematerialization and optimal hybrid parallelism","volume-title":"in 2024 USENIX Annual Technical Conference (USENIX ATC 24)","author":"Yuan"},{"key":"ref59","article-title":"Autohete: An automatic and efficient heterogeneous training system for llms","author":"Zeng","year":"2025","journal-title":"arXiv preprint"},{"key":"ref60","doi-asserted-by":"publisher","DOI":"10.1038\/s41586-025-08628-5"},{"key":"ref61","first-page":"961","article-title":"SmartMoE: Efficiently training Sparsely-Activated models through combining offline and online parallelization","volume-title":"in 2023 USENIX Annual Technical Conference (USENIX ATC 23)","author":"Zhai"},{"key":"ref62","doi-asserted-by":"publisher","DOI":"10.1016\/j.neucom.2023.126661"},{"key":"ref63","article-title":"COMET: Fine-grained computation-communication overlapping for mixture-of-experts","volume-title":"in Eighth Conference on Machine Learning and Systems","author":"Zhang"},{"key":"ref64","doi-asserted-by":"publisher","DOI":"10.1038\/s41586-023-06184-4"},{"key":"ref65","first-page":"559","article-title":"Alpa: Automating inter- and Intra-Operator parallelism for distributed deep learning","volume-title":"in 16th USENIX Symposium on Operating Systems Design and Implementation (OSDI 22)","author":"Zheng"}],"event":{"name":"2026 IEEE International Symposium on High Performance Computer Architecture (HPCA)","location":"Sydney, Australia","start":{"date-parts":[[2026,1,31]]},"end":{"date-parts":[[2026,2,4]]}},"container-title":["2026 IEEE International Symposium on High Performance Computer Architecture (HPCA)"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/11408404\/11408433\/11408533.pdf?arnumber=11408533","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,3,5]],"date-time":"2026-03-05T06:35:58Z","timestamp":1772692558000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11408533\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,1,31]]},"references-count":65,"URL":"https:\/\/doi.org\/10.1109\/hpca68181.2026.11408533","relation":{},"subject":[],"published":{"date-parts":[[2026,1,31]]}}}