{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,1,21]],"date-time":"2026-01-21T22:04:19Z","timestamp":1769033059915,"version":"3.49.0"},"reference-count":15,"publisher":"IEEE","license":[{"start":{"date-parts":[[2025,11,14]],"date-time":"2025-11-14T00:00:00Z","timestamp":1763078400000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2025,11,14]],"date-time":"2025-11-14T00:00:00Z","timestamp":1763078400000},"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":[[2025,11,14]]},"DOI":"10.1109\/cloudcom67567.2025.11331464","type":"proceedings-article","created":{"date-parts":[[2026,1,20]],"date-time":"2026-01-20T20:37:16Z","timestamp":1768941436000},"page":"1-8","source":"Crossref","is-referenced-by-count":0,"title":["System-Aware LLM Arena: Benchmarking Utility, Latency, and Cost"],"prefix":"10.1109","author":[{"given":"Renwen","family":"Ma","sequence":"first","affiliation":[{"name":"Shenzhen Polytechnic University,Shenzhen,China"}]},{"given":"Songrui","family":"Wang","sequence":"additional","affiliation":[{"name":"The Chinese University of Hong Kong, Shenzhen,Shenzhen,China"}]},{"given":"Yicheng","family":"Yu","sequence":"additional","affiliation":[{"name":"Zhejiang University of Technology,Hangzhou,China"}]},{"given":"Tongkai","family":"Ji","sequence":"additional","affiliation":[{"name":"Zhejiang University of Technology,Hangzhou,China"}]},{"given":"Wei","family":"Dai","sequence":"additional","affiliation":[{"name":"Zhejiang University of Technology,Hangzhou,China"}]}],"member":"263","reference":[{"key":"ref1","article-title":"Chatbot arena: An open platform for evaluating llms by human preference","volume-title":"Forty-first International Conference on Machine Learning","author":"Chiang","year":"2024"},{"key":"ref2","article-title":"Mtbench: A multimodal time series benchmark for temporal reasoning and question answering","volume-title":"ar Xiv preprint ar Xiv: 2503.16858","author":"Chen","year":"2025"},{"key":"ref3","author":"Li","year":"2024","journal-title":"From crowdsourced data to high-quality benchmarks: Arena-hard and benchbuilder pipeline"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.1145\/2408776.2408794"},{"key":"ref5","author":"Dao","year":"2023","journal-title":"Flashattention-2: Faster attention with better paral-lelism and work partitioning"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.1145\/3600006.3613165"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v34i09.7123"},{"key":"ref8","author":"Patterson","year":"2021","journal-title":"Carbon emissions and large neural network training"},{"key":"ref9","volume-title":"Holistic evaluation of language models","author":"Liang","year":"2022"},{"key":"ref10","author":"Dettmers","year":"2021","journal-title":"8-bit optimizers via block-wise quantization"},{"key":"ref11","author":"Hendrycks","journal-title":"Measuring massive multitask language understanding"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.findings-naacl.149"},{"key":"ref13","volume-title":"The Rating of Chessplayers, Past and Present","author":"Elo","year":"1978"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.2307\/2334029"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.7551\/mitpress\/7503.003.0076"}],"event":{"name":"2025 lEEE International Conference on Cloud Computing Technology and Science (CloudCom)","location":"Shenzhen, China","start":{"date-parts":[[2025,11,14]]},"end":{"date-parts":[[2025,11,16]]}},"container-title":["2025 lEEE International Conference on Cloud Computing Technology and Science (CloudCom)"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/11330195\/11331311\/11331464.pdf?arnumber=11331464","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,1,21]],"date-time":"2026-01-21T07:55:45Z","timestamp":1768982145000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11331464\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,11,14]]},"references-count":15,"URL":"https:\/\/doi.org\/10.1109\/cloudcom67567.2025.11331464","relation":{},"subject":[],"published":{"date-parts":[[2025,11,14]]}}}