{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,1]],"date-time":"2026-08-01T05:58:14Z","timestamp":1785563894072,"version":"3.56.0"},"reference-count":31,"publisher":"IEEE","license":[{"start":{"date-parts":[[2026,5,1]],"date-time":"2026-05-01T00:00:00Z","timestamp":1777593600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2026,5,1]],"date-time":"2026-05-01T00:00:00Z","timestamp":1777593600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2026,5,1]],"date-time":"2026-05-01T00:00:00Z","timestamp":1777593600000},"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":[[2026,5]]},"DOI":"10.1109\/ccgrid68966.2026.00023","type":"proceedings-article","created":{"date-parts":[[2026,7,31]],"date-time":"2026-07-31T18:07:12Z","timestamp":1785521232000},"page":"136-146","source":"Crossref","is-referenced-by-count":0,"title":["LLM-Pilot: SLO-Aware and Cost-Efficient LLM Serving on Public Cloud VM Clusters via Offloading"],"prefix":"10.1109","author":[{"given":"Jinwoo","family":"Kim","sequence":"first","affiliation":[{"name":"Sogang University"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Kihyun","family":"Kim","sequence":"additional","affiliation":[{"name":"Sogang University"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Hyunsun","family":"Chung","sequence":"additional","affiliation":[{"name":"Sogang University"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jihoon","family":"Yang","sequence":"additional","affiliation":[{"name":"Sogang University"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"James J.","family":"Kim","sequence":"additional","affiliation":[{"name":"Soteria Inc."}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Dong","family":"Li","sequence":"additional","affiliation":[{"name":"University of California","place":["Merced"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Youngjae","family":"Kim","sequence":"additional","affiliation":[{"name":"Sogang University"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","article-title":"Gpt-4 technical report","author":"OpenAI","year":"2024","journal-title":"arXiv preprint"},{"key":"ref2","volume-title":"OpenAI Platform API.","year":"2025"},{"key":"ref3","volume-title":"OpenAI Pricing.","year":"2025"},{"key":"ref4","volume-title":"Message Batches API: Run jobs up to 50% cheaper.","year":"2025"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1177\/1529100615623267"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.1145\/3672198.3673797"},{"key":"ref7","volume-title":"Amazon EC2 Instance Types.","year":"2025"},{"key":"ref8","article-title":"Longrope: Extending 1lm context window beyond 2 million tokens","author":"Ding","year":"2024","journal-title":"arXiv preprint"},{"key":"ref9","article-title":"Llama: Open and efficient foundation language models","author":"Touvron","year":"2023","journal-title":"arXiv preprint"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.1109\/ICDCS63083.2025.00062"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1109\/CCGrid59990.2024.00034"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.1109\/cloud67622.2025.00027"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1145\/3064176.3064182"},{"key":"ref14","article-title":"M\u00e9lange: Cost efficient large language model serving by exploiting gpu heterogeneity","author":"Griggs","year":"2024","journal-title":"arXiv preprint"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.48550\/ARXIV.1706.03762"},{"key":"ref16","first-page":"117","article-title":"Taming throughput-latency tradeoff in LLM inference with Sarathi-Serve","volume-title":"Proceedings of the 18th USENIX Symposium on Operating Systems Design and Implementation (OSDI)","author":"Agrawal","year":"2024"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1145\/3600006.3613165"},{"key":"ref18","volume-title":"Gemini 3.0 pro Context Window Size.","year":"2025"},{"key":"ref19","article-title":"Qwen2.5\u20131m technical report","author":"Yang","year":"2025","journal-title":"arXiv preprint"},{"key":"ref20","article-title":"Long context rag performance of large language models","author":"Leng","year":"2024","journal-title":"arXiv preprint"},{"key":"ref21","first-page":"103","article-title":"GPipe: Efficient training of giant neural networks using pipeline parallelism","volume-title":"Proceedings of the33rd International Conference on Neural Information Processing Systems (NeurIPS)","author":"Huang"},{"key":"ref22","article-title":"Megatron-Im: Training multi-billion parameter language models using model parallelism","author":"Shoeybi","year":"2019","journal-title":"arXiv preprint"},{"key":"ref23","article-title":"FlexGen: High-throughput generative inference of large language models with a single GPU","volume-title":"Proceedings of the 40th International Conference on Machine Learning (ICML), JMLR.org","author":"Sheng","year":"2023"},{"key":"ref24","author":"Xiong","year":"2024","journal-title":"LayerKV: Optimizing large language model serving with layer-wise KV cache management"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.1109\/SC41404.2022.00051"},{"key":"ref26","article-title":"Instinfer: In-storage attention offloading for cost-effective long-context llm inference","author":"Pan","year":"2024","journal-title":"arXiv preprint"},{"key":"ref27","doi-asserted-by":"publisher","DOI":"10.1007\/s00607-023-01173-x"},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.1109\/ISCA59077.2024.00019"},{"key":"ref29","article-title":"Theory, Volume 1","volume-title":"Queueing Systems","author":"Kleinrock","year":"1975"},{"key":"ref30","volume-title":"ShareGPT: Share your wildest ChatGPT conversations.","year":"2025"},{"key":"ref31","volume-title":"Stanford Alpaca: An Instruction-following LLaMA Model.","author":"Taori","year":"2025"}],"event":{"name":"2026 IEEE 26th International Symposium on Cluster, Cloud and Internet Computing (CCGrid)","location":"Sydney, Australia","start":{"date-parts":[[2026,5,18]]},"end":{"date-parts":[[2026,5,21]]}},"container-title":["2026 IEEE 26th International Symposium on Cluster, Cloud and Internet Computing (CCGrid)"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/11618979\/11618983\/11619230.pdf?arnumber=11619230","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,1]],"date-time":"2026-08-01T05:11:16Z","timestamp":1785561076000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11619230\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,5]]},"references-count":31,"URL":"https:\/\/doi.org\/10.1109\/ccgrid68966.2026.00023","relation":{},"subject":[],"published":{"date-parts":[[2026,5]]}}}