{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,21]],"date-time":"2026-05-21T20:14:05Z","timestamp":1779394445305,"version":"3.53.1"},"reference-count":39,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","issue":"6","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"}],"funder":[{"DOI":"10.13039\/501100012166","name":"National Key Research and Development Program of China","doi-asserted-by":"publisher","award":["2022YFB4501600"],"award-info":[{"award-number":["2022YFB4501600"]}],"id":[{"id":"10.13039\/501100012166","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100008100","name":"Youth Innovation Fund of the School of Software Engineering, University of Science and Technology of China","doi-asserted-by":"publisher","award":["YN2260080008"],"award-info":[{"award-number":["YN2260080008"]}],"id":[{"id":"10.13039\/501100008100","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE Trans. Comput.-Aided Des. Integr. Circuits Syst."],"published-print":{"date-parts":[[2026,6]]},"DOI":"10.1109\/tcad.2025.3624135","type":"journal-article","created":{"date-parts":[[2025,10,21]],"date-time":"2025-10-21T17:09:17Z","timestamp":1761066557000},"page":"2826-2839","source":"Crossref","is-referenced-by-count":0,"title":["AsyncGrid: An Intralayer and Interlayer Asynchronous Hybrid Parallelism System for Responsive Edge LLM Inference"],"prefix":"10.1109","volume":"45","author":[{"ORCID":"https:\/\/orcid.org\/0009-0002-7439-2505","authenticated-orcid":false,"given":"Yi","family":"Xiong","sequence":"first","affiliation":[{"name":"School of Computer Science and Technology, University of Science and Technology of China, Hefei, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Rui","family":"Zhang","sequence":"additional","affiliation":[{"name":"Suzhou Institute for Advanced Research, University of Science and Technology of China, Suzhou, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yulong","family":"Zu","sequence":"additional","affiliation":[{"name":"Suzhou Institute for Advanced Research, University of Science and Technology of China, Suzhou, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-0308-0259","authenticated-orcid":false,"given":"Weihong","family":"Liu","sequence":"additional","affiliation":[{"name":"School of Computer Science and Technology, University of Science and Technology of China, Hefei, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-3607-2631","authenticated-orcid":false,"given":"Zongwei","family":"Zhu","sequence":"additional","affiliation":[{"name":"Suzhou Institute for Advanced Research, University of Science and Technology of China, Suzhou, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-7405-9820","authenticated-orcid":false,"given":"Jiawei","family":"Geng","sequence":"additional","affiliation":[{"name":"School of Computer Science and Technology, University of Science and Technology of China, Hefei, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0001-2796-2686","authenticated-orcid":false,"given":"Boyu","family":"Li","sequence":"additional","affiliation":[{"name":"School of Computer Science and Technology, University of Science and Technology of China, Hefei, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0001-2861-8211","authenticated-orcid":false,"given":"Qianyue","family":"Cao","sequence":"additional","affiliation":[{"name":"School of Computer Science and Technology, University of Science and Technology of China, Hefei, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-8360-3143","authenticated-orcid":false,"given":"Xuehai","family":"Zhou","sequence":"additional","affiliation":[{"name":"School of Computer Science and Technology, University of Science and Technology of China, Hefei, China"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","article-title":"Qwen2 technical report","volume-title":"arXiv:2407.10671","author":"Yang","year":"2024"},{"key":"ref2","article-title":"Phi-3 technical report: A highly capable language model locally on your phone","volume-title":"arXiv:2404. 14219","author":"Abdin","year":"2024"},{"key":"ref3","article-title":"GPT-4 technical report","volume-title":"arXiv:2303. 08774","author":"Achiam","year":"2023"},{"key":"ref4","article-title":"LayerKV: Optimizing large language model serving with layer-wise KV cache management","author":"Xiong","year":"2024","journal-title":"arXiv:2410.00428"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1145\/1476589.1476628"},{"key":"ref6","article-title":"Megatron-LM: Training multi-billion parameter language models using model parallelism","author":"Shoeybi","year":"2019","journal-title":"arXiv:1909.08053"},{"key":"ref7","article-title":"Ring attention with blockwise transformers for near-infinite context","author":"Liu","year":"2023","journal-title":"arXiv:2310.01889"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1145\/3341301.3359646"},{"key":"ref9","first-page":"173","article-title":"Llumnix: Dynamic scheduling for large language model serving","volume-title":"Proc. 18th USENIX Symp. Operating Syst. Design Implement.","author":"Sun"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.1016\/j.jml.2019.104047"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1109\/TCAD.2024.3513892"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.1109\/TSC.2025.3552338"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1109\/TMC.2025.3553201"},{"key":"ref14","article-title":"Fast on-device LLM inference with NPUs","author":"Xu","year":"2024","journal-title":"arXiv:2407.05858"},{"key":"ref15","article-title":"BurstGPT: A real-world workload dataset to optimize LLM serving systems","author":"Wang","year":"2024","journal-title":"arXiv:2401.17644"},{"key":"ref16","article-title":"DeepSpeed ulysses: System optimizations for enabling training of extreme long sequence transformer models","author":"Jacobs","year":"2023","journal-title":"arXiv:2309.14509"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2023.acl-long.134"},{"key":"ref18","article-title":"Reducing activation recomputation in large transformer models","volume-title":"Proc. 6th Conf. Mach. Learn. Syst. (MLSys)","author":"Korthikanti"},{"key":"ref19","article-title":"Striped attention: Faster ring attention for causal transformers","author":"Brandon","year":"2023","journal-title":"arXiv:2311.09431"},{"key":"ref20","first-page":"6543","article-title":"TeraPipe: Token-level pipeline parallelism for training large-scale language models","volume-title":"Proc. 38th Int. Conf. Mach. Learn.","volume":"139","author":"Li"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.1145\/3591300"},{"key":"ref22","first-page":"193","article-title":"DistServe: Disaggregating prefill and decoding for goodput-optimized large language model serving","volume-title":"Proc. 18th USENIX Symp. Operating Syst. Design Implement.","author":"Zhong"},{"key":"ref23","first-page":"31094","article-title":"FlexGen: High-throughput generative inference of large language models with a single GPU","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Sheng"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.1016\/j.procs.2010.04.241"},{"key":"ref25","volume-title":"CP-SAT Solver\u2014Constraint Programming in OR-Tools","year":"2025"},{"key":"ref26","first-page":"663","article-title":"AlpaServe: Statistical multiplexing with model parallelism for deep learning serving","volume-title":"Proc. 17th USENIX Symp. Operating Syst. Design Implement.","author":"Li"},{"key":"ref27","volume-title":"Deep Learning Performance\u2014Matrix Multiplication Optimization","year":"2025"},{"key":"ref28","volume-title":"ShareGPT: Share and Explore ChatGPT Conversations","year":"2023"},{"key":"ref29","doi-asserted-by":"publisher","DOI":"10.1145\/3600006.3613165"},{"key":"ref30","article-title":"MInference 1.0: Accelerating pre-filling for long-context LLMs via dynamic sparse attention","author":"Jiang","year":"2024","journal-title":"arXiv:2407.02490"},{"key":"ref31","article-title":"GPTQ: Accurate post-training quantization for generative pre-trained transformers","author":"Frantar","year":"2022","journal-title":"arXiv:2210.17323"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2023.emnlp-main.298"},{"key":"ref33","article-title":"GShard: Scaling giant models with conditional computation and automatic sharding","volume-title":"Proc. 9th Int. Conf. Learn. Represent. (ICLR)","author":"Lepikhin"},{"key":"ref34","article-title":"SLEB: Streamlining LLMs through redundancy verification and elimination of transformer blocks","volume-title":"Proc. 41st Int. Conf. Mach. Learn.","author":"Song"},{"key":"ref35","article-title":"FlashAttention-2: Faster attention with better parallelism and work partitioning","volume-title":"Proc. 12th Int. Conf. Learn. Represent. (ICLR)","author":"Dao"},{"key":"ref36","article-title":"Fast distributed inference serving for large language models","author":"Wu","year":"2023","journal-title":"arXiv:2305.05920"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.1145\/3773772"},{"key":"ref38","doi-asserted-by":"publisher","DOI":"10.1109\/ISCA59077.2024.00019"},{"key":"ref39","article-title":"Efficiently scaling transformer inference","volume-title":"Proc. 6th Conf. Mach. Learn. Syst. (MLSys)","author":"Pope"}],"container-title":["IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/43\/11527413\/11214161.pdf?arnumber=11214161","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,5,21]],"date-time":"2026-05-21T19:43:12Z","timestamp":1779392592000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11214161\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,6]]},"references-count":39,"journal-issue":{"issue":"6"},"URL":"https:\/\/doi.org\/10.1109\/tcad.2025.3624135","relation":{},"ISSN":["0278-0070","1937-4151"],"issn-type":[{"value":"0278-0070","type":"print"},{"value":"1937-4151","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,6]]}}}