{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,26]],"date-time":"2026-05-26T15:04:22Z","timestamp":1779807862288,"version":"3.53.1"},"publisher-location":"New York, NY, USA","reference-count":10,"publisher":"ACM","funder":[{"name":"NSERC","award":["RGPIN-2023-03478"],"award-info":[{"award-number":["RGPIN-2023-03478"]}]},{"name":"NSF","award":["CCF-2413870"],"award-info":[{"award-number":["CCF-2413870"]}]}],"content-domain":{"domain":["dl.acm.org"],"crossmark-restriction":true},"short-container-title":[],"published-print":{"date-parts":[[2026,6,8]]},"DOI":"10.1145\/3801489.3806850","type":"proceedings-article","created":{"date-parts":[[2026,5,26]],"date-time":"2026-05-26T14:22:48Z","timestamp":1779805368000},"page":"94-96","update-policy":"https:\/\/doi.org\/10.1145\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["Cache Your Prompt When It's Green - Carbon-aware Caching for Large Language Model Serving"],"prefix":"10.1145","author":[{"ORCID":"https:\/\/orcid.org\/0009-0006-5438-1795","authenticated-orcid":false,"given":"Yuyang","family":"Tian","sequence":"first","affiliation":[{"name":"University of Waterloo, Waterloo, Canada"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-8630-7959","authenticated-orcid":false,"given":"Desen","family":"Sun","sequence":"additional","affiliation":[{"name":"University of Waterloo, Waterloo, Canada"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-2757-9182","authenticated-orcid":false,"given":"Yi","family":"Ding","sequence":"additional","affiliation":[{"name":"Purdue University, West Lafayette, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-9706-6177","authenticated-orcid":false,"given":"Sihang","family":"Liu","sequence":"additional","affiliation":[{"name":"University of Waterloo, Waterloo, Canada"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"320","published-online":{"date-parts":[[2026,6,8]]},"reference":[{"key":"e_1_3_2_1_1_1","volume-title":"Azure LLM inference trace","author":"Azure","year":"2024","unstructured":"Azure. 2024. Azure LLM inference trace 2024. https:\/\/github.com\/Azure\/AzurePublicDataset\/tree\/master."},{"key":"e_1_3_2_1_2_1","unstructured":"Ahmad Faiz et al. 2024. LLMCarbon: Modeling the End-to-End Carbon Footprint of Large Language Models. In ICLR. https:\/\/openreview.net\/forum?id=aIok3ZD9to"},{"key":"e_1_3_2_1_3_1","doi-asserted-by":"publisher","unstructured":"Udit Gupta and other. 2022. ACT: designing sustainable computer systems with an architectural carbon modeling tool. In ISCA. doi:10.1145\/3470496.3527408","DOI":"10.1145\/3470496.3527408"},{"key":"e_1_3_2_1_4_1","doi-asserted-by":"publisher","unstructured":"Woosuk Kwon et al. 2023. Efficient Memory Management for Large Language Model Serving with PagedAttention. In SOSP. doi:10.1145\/3600006.3613165","DOI":"10.1145\/3600006.3613165"},{"key":"e_1_3_2_1_5_1","doi-asserted-by":"publisher","unstructured":"Yuhan Liu et al. 2024. CacheGen: KV Cache Compression and Streaming for Fast Large Language Model Serving. In SIGCOMM. doi:10.1145\/3651890.3672274","DOI":"10.1145\/3651890.3672274"},{"key":"e_1_3_2_1_6_1","unstructured":"LMCache Team. 2025. LMCache. https:\/\/lmcache.ai\/."},{"key":"e_1_3_2_1_7_1","unstructured":"Diptyaroop Maji et al. 2023. Multi-Day Forecasting of Electric Grid Carbon Intensity Using Machine Learning. In BuildSys."},{"key":"e_1_3_2_1_8_1","unstructured":"ShareGPT. 2023. https:\/\/sharegpt.com\/."},{"key":"e_1_3_2_1_9_1","unstructured":"Tianyao Shi et al. 2024. GreenLLM: Disaggregating Large Language Model Serving on Heterogeneous GPUs for Lower Carbon Emissions. arXiv:2412.20322 [cs.AR] https:\/\/arxiv.org\/abs\/2412.20322"},{"key":"e_1_3_2_1_10_1","doi-asserted-by":"publisher","unstructured":"Swamit Tannu et al. 2023. The Dirty Secret of SSDs: Embodied Carbon. SIGENERGY Energy Inform. Rev. (2023). doi:10.1145\/3630614.3630616","DOI":"10.1145\/3630614.3630616"}],"event":{"name":"SIGMETRICS '26: ACM SIGMETRICS International Conference on Measurement and Modeling of Computer Systems","location":"Ann Arbor MI USA","sponsor":["SIGMETRICS ACM Special Interest Group on Measurement and Evaluation"]},"container-title":["Abstracts of the 2026 ACM SIGMETRICS International Conference on Measurement and Modeling of Computer Systems"],"original-title":[],"deposited":{"date-parts":[[2026,5,26]],"date-time":"2026-05-26T14:26:13Z","timestamp":1779805573000},"score":1,"resource":{"primary":{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3801489.3806850"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,6,8]]},"references-count":10,"alternative-id":["10.1145\/3801489.3806850","10.1145\/3801489"],"URL":"https:\/\/doi.org\/10.1145\/3801489.3806850","relation":{},"subject":[],"published":{"date-parts":[[2026,6,8]]},"assertion":[{"value":"2026-06-08","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}