{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,11,24]],"date-time":"2025-11-24T19:55:40Z","timestamp":1764014140646,"version":"3.45.0"},"reference-count":12,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","issue":"2","license":[{"start":{"date-parts":[[2025,7,1]],"date-time":"2025-07-01T00:00:00Z","timestamp":1751328000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2025,7,1]],"date-time":"2025-07-01T00:00:00Z","timestamp":1751328000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2025,7,1]],"date-time":"2025-07-01T00:00:00Z","timestamp":1751328000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"name":"Institute of Information & Communications Technology Planning","award":["RS-2022-II221037","RS-2024-00396013"],"award-info":[{"award-number":["RS-2022-II221037","RS-2024-00396013"]}]},{"DOI":"10.13039\/501100003725","name":"National Research Foundation of Korea","doi-asserted-by":"publisher","award":["RS-2024-00342148"],"award-info":[{"award-number":["RS-2024-00342148"]}],"id":[{"id":"10.13039\/501100003725","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100003696","name":"Electronics and Telecommunications Research Institute","doi-asserted-by":"publisher","award":["RS-2025-02305453"],"award-info":[{"award-number":["RS-2025-02305453"]}],"id":[{"id":"10.13039\/501100003696","id-type":"DOI","asserted-by":"publisher"}]},{"name":"Korea Government"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE Comput. Arch. Lett."],"published-print":{"date-parts":[[2025,7]]},"DOI":"10.1109\/lca.2025.3628325","type":"journal-article","created":{"date-parts":[[2025,11,3]],"date-time":"2025-11-03T18:43:53Z","timestamp":1762195433000},"page":"361-364","source":"Crossref","is-referenced-by-count":0,"title":["LLMServingSim2.0: A Unified Simulator for Heterogeneous Hardware and Serving Techniques in LLM Infrastructure"],"prefix":"10.1109","volume":"24","author":[{"ORCID":"https:\/\/orcid.org\/0009-0008-8973-5164","authenticated-orcid":false,"given":"Jaehong","family":"Cho","sequence":"first","affiliation":[{"name":"School of Computing, Korea Advanced Institute of Science and Technology, Daejeon, South Korea"}]},{"ORCID":"https:\/\/orcid.org\/0009-0002-4116-3896","authenticated-orcid":false,"given":"Hyunmin","family":"Choi","sequence":"additional","affiliation":[{"name":"School of Computing, Korea Advanced Institute of Science and Technology, Daejeon, South Korea"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-6629-449X","authenticated-orcid":false,"given":"Jongse","family":"Park","sequence":"additional","affiliation":[{"name":"School of Computing, Korea Advanced Institute of Science and Technology, Daejeon, South Korea"}]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.1109\/IISWC63097.2024.00012"},{"key":"ref2","first-page":"351","article-title":"Vidur: A large-scale simulation framework for LLM inference","volume-title":"Proc. of ML. and Sys.","author":"Agrawal","year":"2024"},{"article-title":"APEX: An extensible and dynamism-aware simulator for automated parallel execution in LLM serving","year":"2025","author":"Lin","key":"ref3"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.1007\/978-981-95-1021-4_19"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1109\/ISCA59077.2024.00019"},{"key":"ref6","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. Des. Implementation","author":"Zhong","year":"2024"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1109\/ISCA59077.2024.00078"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1109\/MICRO61859.2024.00105"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1145\/3600006.3613165"},{"key":"ref10","first-page":"62557","article-title":"SGLang: Efficient execution of structured language model programs","volume-title":"Adv. in Neural Info. Proc. Sys.","author":"Zheng","year":"2024"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1145\/3689031.3696098"},{"year":"2023","key":"ref12","article-title":"ShareGPT"}],"container-title":["IEEE Computer Architecture Letters"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/10208\/11062520\/11224567.pdf?arnumber=11224567","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,11,24]],"date-time":"2025-11-24T18:59:09Z","timestamp":1764010749000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11224567\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,7]]},"references-count":12,"journal-issue":{"issue":"2"},"URL":"https:\/\/doi.org\/10.1109\/lca.2025.3628325","relation":{},"ISSN":["1556-6056","1556-6064","2473-2575"],"issn-type":[{"type":"print","value":"1556-6056"},{"type":"electronic","value":"1556-6064"},{"type":"electronic","value":"2473-2575"}],"subject":[],"published":{"date-parts":[[2025,7]]}}}