{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,5]],"date-time":"2026-08-05T06:44:59Z","timestamp":1785912299589,"version":"3.56.0"},"reference-count":80,"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\/isca66397.2026.00077","type":"proceedings-article","created":{"date-parts":[[2026,8,4]],"date-time":"2026-08-04T19:17:09Z","timestamp":1785871029000},"page":"969-984","source":"Crossref","is-referenced-by-count":0,"title":["DynoPipe: Heterogeneous Edge-Cloud LLM Serving with Dynamically Orchestrated Pipeline Boundaries"],"prefix":"10.1109","author":[{"given":"Yanying","family":"Lin","sequence":"first","affiliation":[{"name":"Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Baicheng","family":"Chen","sequence":"additional","affiliation":[{"name":"University of California San Diego"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xinyu","family":"Zhang","sequence":"additional","affiliation":[{"name":"University of California San Diego"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Chengzhong","family":"Xu","sequence":"additional","affiliation":[{"name":"University of Macau"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Kejiang","family":"Ye","sequence":"additional","affiliation":[{"name":"Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.1109\/isca59077.2024.00049"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.1145\/3470496.3527382"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.1145\/3575693.3575745"},{"key":"ref4","first-page":"209","article-title":"Privid: Practical, privacy-preserving video analytics queries","author":"Cangialosi","year":"2022","journal-title":"19th USENIX Symp. Networked Syst. Des. Implement. NSDI 2022 Renton WA USA April 4-6 2022, ser. NSDI 2022"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1145\/3341301.3359639"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.1109\/ACCESS.2020.3045078"},{"key":"ref7","first-page":"117","article-title":"Taming throughputlatency tradeoff in llm inference with sarathi-serve","volume-title":"18th USENIX Symp. Oper. Syst. Des. Implement. OSDI 2024 St. Clara CA USA July 10\u201312 2024","author":"Agrawal","year":"2024"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1109\/ICDCS60910.2024.00010"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1145\/3466752.3480068"},{"key":"ref10","first-page":"1629","article-title":"Depth gives a false sense of privacy:{LLM} internal states inversion","volume-title":"34th USENIX Security Symposium (USENIX Security 25)","author":"Dong","year":"2025"},{"key":"ref11","author":"Kim","year":"2024","journal-title":"Squeezellm: Dense-and-sparse quantization"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.1145\/3466752.3480123"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2020.emnlp-main.633"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.1145\/3297858.3304012"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.1145\/3552326.3567485"},{"key":"ref16","first-page":"1029","article-title":"Efficient decentralized federated singular vector decomposition","volume-title":"Proc. 2024 USENIX Annu. Tech. Conf. USENIX ATC 2024","author":"Chai"},{"key":"ref17","first-page":"579","article-title":"Fwdllm: Efficient federated finetuning of large language models with perturbed inferences","volume-title":"Proc. 2024 USENIX Annu. Tech. Conf. USENIX ATC 2024","author":"Xu"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1109\/isca59077.2024.00081"},{"key":"ref19","first-page":"199","article-title":"Serving heterogeneous machine learning models on multi-gpu servers with spatio-temporal sharing","volume-title":"Proc. 2022 USENIX Annu. Tech. Conf. USENIX ATC 2022","author":"Choi"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.1145\/2556583"},{"key":"ref21","first-page":"601","article-title":"Overcoming the memory wall with cxlenabled ssds","volume-title":"Proc. 2023 USENIX Annu. Tech. Conf. USENIX ATC 2023 Boston MA USA July 10-12 2023, ser. USENIX ATC 2023","author":"Yang","year":"2023"},{"key":"ref22","first-page":"409","article-title":"Taming performance variability","volume-title":"13th USENIX Symp. Oper. Syst. Des. Implement. OSDI 2018","author":"Maricq"},{"key":"ref23","first-page":"708","article-title":"Netren: Service migration-driven network renascence with synthesizing upyeard configuration","volume-title":"Proc. 29th ACM Int. Conf. Archit. Support Program. Lang. Oper. Syst.","volume":"3","author":"Han","year":"2024\u201304-27"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.1145\/3575693.3575736"},{"key":"ref25","first-page":"47","article-title":"Migros: Transparent live-migration support for containerised rdma applications","volume-title":"Proc. 2021 USENIX Annu. Tech. Conf. USENIX ATC 2021 July 14-16 2021, ser. USENIX ATC 2021","author":"Planeta","year":"2021"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1145\/3620666.3651359"},{"key":"ref27","article-title":"Edgeshard: Efficient llm inference via collaborative edge computing","author":"Zhang","year":"2024","journal-title":"arXiv.org"},{"key":"ref28","author":"Hu","year":"2025","journal-title":"Efficient long-decoding inference with reasoning-aware attention sparsity"},{"key":"ref29","author":"Yue","year":"2025","journal-title":"Inference scaling for long-context retrieval augmented generation"},{"key":"ref30","doi-asserted-by":"publisher","DOI":"10.1109\/HPCA51647.2021.00013"},{"key":"ref31","doi-asserted-by":"publisher","DOI":"10.1016\/j.eswa.2026.131632"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.1109\/ICCAD66269.2025.11240746"},{"key":"ref33","article-title":"Privacypreserving llm inference in practice: A comparative survey of techniques, trade-offs, and deployability","author":"Andreoletti","year":"2026","journal-title":"Cryptology ePrint Archive"},{"key":"ref34","author":"Gao","year":"2025","journal-title":"Seerattention: Learning intrinsic sparse attention in your llms"},{"key":"ref35","doi-asserted-by":"publisher","DOI":"10.52202\/079017-1663"},{"key":"ref36","author":"Xiao","year":"2024","journal-title":"Smoothquant: Accurate and efficient post-training quantization for large language models"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.1145\/3477132.3483580"},{"key":"ref38","first-page":"787","article-title":"Shepherd: Serving dnns in the wild","volume-title":"20th USENIX Symp. Networked Syst. Des. Implement. NSDI 2023","author":"Zhang"},{"key":"ref39","doi-asserted-by":"publisher","DOI":"10.1145\/3620665.3640411"},{"key":"ref40","first-page":"911","article-title":"dlora: Dynamically orchestrating requests and adapters for lora llm serving","volume-title":"18th USENIX Symp. Oper. Syst. Des. Implement. OSDI 2024","author":"Wu"},{"key":"ref41","first-page":"663","article-title":"Alpaserve: Statistical multiplexing with model parallelism for deep learning serving","volume-title":"17th USENIX Symp. Oper. Syst. Des. Implement. OSDI 2023","author":"Li"},{"key":"ref42","doi-asserted-by":"publisher","DOI":"10.1109\/isca59077.2024.00019"},{"key":"ref43","doi-asserted-by":"publisher","DOI":"10.1145\/3636534.3649391"},{"key":"ref44","article-title":"Key metrics for LLM inference","year":"2025","journal-title":"LLM Inference Handbook"},{"key":"ref45","article-title":"Ab - apache http server benchmarking tool - apache http server version 2.4","journal-title":"Apache"},{"key":"ref46","doi-asserted-by":"publisher","DOI":"10.1145\/3093337.3037698"},{"key":"ref47","doi-asserted-by":"publisher","DOI":"10.1145\/3267809.3267828"},{"key":"ref48","doi-asserted-by":"publisher","DOI":"10.1109\/INFCOMW.2019.8845240"},{"key":"ref49","doi-asserted-by":"publisher","DOI":"10.1109\/TCAD.2018.2858384"},{"key":"ref50","doi-asserted-by":"publisher","DOI":"10.1109\/ICDCS47774.2020.00101"},{"key":"ref51","doi-asserted-by":"publisher","DOI":"10.1145\/3397315"},{"key":"ref52","first-page":"551","article-title":"Zero-offload: Democratizing billionscale model training","volume-title":"Proc. 2021 USENIX Annu. Tech. Conf. USENIX ATC","author":"Ren","year":"2021 July 14-16 2021"},{"key":"ref53","article-title":"Tango: Secure collaborative route control across the public internet","volume-title":"21st USENIX Symp. Networked Syst. Des. Implement. NSDI 2024 St. Clara CA April 15-17 2024","author":"Birge-Lee","year":"2024"},{"key":"ref54","first-page":"1031","article-title":"Secrecy: Secure collaborative analytics in untrusted clouds","author":"Liagouris","year":"2023","journal-title":"20th USENIX Symp. Networked Syst. Des. Implement. NSDI 2023 Boston MA April 17-19 2023, ser. NSDI 2023. USENIX Association"},{"key":"ref55","doi-asserted-by":"publisher","DOI":"10.1109\/icws67624.2025.00046"},{"key":"ref56","first-page":"559","article-title":"Alpa: Automating inter- and intra-operator parallelism for distributed deep learning","author":"Zheng","year":"2022","journal-title":"16th USENIX Symp. Oper. Syst. Des. Implement. OSDI 2022 Carlsbad CA USA July 11-13 2022, ser. OSDI 2022"},{"key":"ref57","doi-asserted-by":"publisher","DOI":"10.1145\/3341301.3359646"},{"key":"ref58","first-page":"7937","article-title":"Memory-efficient pipeline-parallel dnn training","volume-title":"Proc. 38th Int. Conf. Mach. Learn. ICML 2021 18-24 July 2021 Virtual Event, ser. ICML 2021. PMLR","author":"Narayanan","year":"2021"},{"key":"ref59","first-page":"499","article-title":"Pipeswitch: Fast pipelined context switching for deep learning applications","volume-title":"14th USENIX Symp. Oper. Syst. Des. Implement. OSDI 2020 Virtual Event Novemb. 4-6 2020, ser. OSDI 2020","author":"Bai","year":"2020"},{"key":"ref60","author":"Shazeer","year":"2017","journal-title":"Outrageously large neural networks: The sparsely-gated mixture-of-experts layer"},{"key":"ref61","doi-asserted-by":"publisher","DOI":"10.1109\/SC41404.2022.00051"},{"key":"ref62","doi-asserted-by":"publisher","DOI":"10.1145\/3437801.3441593"},{"key":"ref63","doi-asserted-by":"publisher","DOI":"10.1145\/3423211.3425692"},{"key":"ref64","first-page":"1089","article-title":"Crosspipe: Towards optimal pipeline schedules for cross-datacenter training","volume-title":"Proc. 2025 USENIX Annu. Tech. Conf. USENIX ATC 2025 Boston MA USA July 7-9 2025","author":"Chen","year":"2025"},{"key":"ref65","doi-asserted-by":"publisher","DOI":"10.1145\/3689031.3717469"},{"key":"ref66","doi-asserted-by":"publisher","DOI":"10.1145\/3669940.3707224"},{"key":"ref67","doi-asserted-by":"publisher","DOI":"10.1109\/asap52443.2021.00020"},{"key":"ref68","doi-asserted-by":"publisher","DOI":"10.1145\/3373087.3375311"},{"key":"ref69","doi-asserted-by":"publisher","DOI":"10.1145\/3575693.3575698"},{"key":"ref70","doi-asserted-by":"publisher","DOI":"10.1109\/JSAC.2020.3005469"},{"key":"ref71","doi-asserted-by":"publisher","DOI":"10.1145\/3318216.3363370"},{"key":"ref72","doi-asserted-by":"publisher","DOI":"10.1109\/ICDCS.2017.94"},{"key":"ref73","author":"Liu","year":"2024","journal-title":"Cachegen: Fast context loading for language model applications via kv cache streaming"},{"key":"ref74","author":"Strati","year":"2024","journal-title":"Dejavu: Kv-cache streaming for fast, fault-tolerant generative llm serving"},{"key":"ref75","doi-asserted-by":"publisher","DOI":"10.1145\/3603269.3604830"},{"key":"ref76","first-page":"119","article-title":"Ekya: Continuous learning of video analytics models on edge compute servers","volume-title":"19th USENIX Symp. Networked Syst. Des. Implement. NSDI 2022 Renton WA USA April 4-6 2022, ser. NSDI 2022. USENIX Association","author":"Bhardwaj","year":"2022"},{"key":"ref77","author":"Lin","year":"2024","journal-title":"Qserve: W4a8kv4 quantization and system co-design for efficient llm serving"},{"key":"ref78","doi-asserted-by":"publisher","DOI":"10.1145\/3617232.3624849"},{"key":"ref79","first-page":"443","article-title":"Serving dnns like clockwork: Performance predictability from the bottom up","volume-title":"14th USENIX Symp. Oper. Syst. Des. Implement. OSDI 2020 Virtual Event Novemb. 4-6 2020, ser. OSDI 2020. USENIX Association","author":"Gujarati","year":"2020"},{"key":"ref80","first-page":"521","article-title":"Orca: A distributed serving system for transformer-based generative models","volume-title":"16th USENIX Symp. Oper. Syst. Des. Implement. OSDI 2022 Carlsbad CA USA July 11-13 2022, ser. OSDI 2022. USENIX Association","author":"Yu","year":"2022"}],"event":{"name":"2026 ACM\/IEEE 53rd Annual International Symposium on Computer Architecture (ISCA)","location":"Raleigh, NC, USA","start":{"date-parts":[[2026,6,27]]},"end":{"date-parts":[[2026,7,1]]}},"container-title":["2026 ACM\/IEEE 53rd Annual International Symposium on Computer Architecture (ISCA)"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/11617393\/11617394\/11617479.pdf?arnumber=11617479","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,5]],"date-time":"2026-08-05T05:51:26Z","timestamp":1785909086000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11617479\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,6]]},"references-count":80,"URL":"https:\/\/doi.org\/10.1109\/isca66397.2026.00077","relation":{},"subject":[],"published":{"date-parts":[[2026,6]]}}}