{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,1]],"date-time":"2026-07-01T08:48:51Z","timestamp":1782895731209,"version":"3.54.5"},"reference-count":42,"publisher":"IEEE","license":[{"start":{"date-parts":[[2026,5,25]],"date-time":"2026-05-25T00:00:00Z","timestamp":1779667200000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2026,5,25]],"date-time":"2026-05-25T00:00:00Z","timestamp":1779667200000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"DOI":"10.13039\/100000001","name":"National Science Foundation","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100000001","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100000015","name":"U.S. Department of Energy","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100000015","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2026,5,25]]},"DOI":"10.1109\/ipdps65963.2026.00083","type":"proceedings-article","created":{"date-parts":[[2026,6,30]],"date-time":"2026-06-30T20:55:16Z","timestamp":1782852916000},"page":"970-982","source":"Crossref","is-referenced-by-count":0,"title":["Beyond Throughput: Performance and Energy Insights of LLM Inference Across AI Accelerators"],"prefix":"10.1109","author":[{"given":"Giacomo","family":"Brunetta","sequence":"first","affiliation":[{"name":"University of Illinois Chicago,Chicago,IL,USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Varuni","family":"Sastry","sequence":"additional","affiliation":[{"name":"Argonne National Laboratory,Lemont,IL,USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xingfu","family":"Wu","sequence":"additional","affiliation":[{"name":"Argonne National Laboratory,Lemont,IL,USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Valerie","family":"Taylor","sequence":"additional","affiliation":[{"name":"Argonne National Laboratory,Lemont,IL,USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Michael E.","family":"Papka","sequence":"additional","affiliation":[{"name":"Argonne National Laboratory,Lemont,IL,USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Zhiling","family":"Lan","sequence":"additional","affiliation":[{"name":"University of Illinois Chicago,Chicago,IL,USA"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","article-title":"Energy and ai: Energy demand from ai","year":"2025"},{"key":"ref2","article-title":"Carbon emissions and large neural network training","author":"Patterson","year":"2021"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.3390\/app15020586"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.1145\/3729215"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2025.acl-long.1563"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.1145\/3721146.3721953"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1145\/3727200.3727217"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1109\/HPCA61900.2025.00102"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1109\/LCA.2024.3406038"},{"key":"ref10","article-title":"Mlperf inference benchmark","author":"Reddi","year":"2019"},{"key":"ref11","doi-asserted-by":"crossref","article-title":"Llm-inference-bench: Inference benchmarking of large language models on ai accelerators","author":"Chitty-Venkata","DOI":"10.1109\/SCW63240.2024.00178"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.1109\/IPDPSW63119.2024.00016"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1109\/PMBS56514.2022.00007"},{"key":"ref14","article-title":"Argonne leadership computing facility (alcf)","journal-title":"U.S. DOE Office of Science User Facility, this research used resources of the Argonne Leadership Computing Facility, which is a DOE Office of Science User Facility supported under Contract DE-AC02-06CH11357"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.1109\/ISSCC42613.2021.9365803"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.1109\/MM.2023.3256796"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1145\/3673038.3673110"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1109\/ISSCC42614.2022.9731673"},{"key":"ref19","doi-asserted-by":"crossref","DOI":"10.1109\/MICRO61859.2024.00100","article-title":"Sambanova sn40l: Scaling the ai memory wall with dataflow and composition of experts","author":"Prabhakar","year":"2024"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.1109\/MM.2023.3256384"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.48550\/ARXIV.1706.03762"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2022.emnlp-main.804"},{"key":"ref23","article-title":"Scaling laws for neural language models","author":"Kaplan","year":"2020"},{"key":"ref24","article-title":"The llama 3 herd of models","author":"Grattafiori","year":"2024"},{"key":"ref25","article-title":"Llama 4: Multimodal intelligence","year":"2025"},{"key":"ref26","article-title":"Qwen2. 5 technical report","author":"Yang","year":"2024"},{"key":"ref27","article-title":"Qwen3 technical report","author":"Yang","year":"2025"},{"key":"ref28","article-title":"Mistral 7b. arxiv 2023","author":"Jiang","year":"2024"},{"key":"ref29","doi-asserted-by":"publisher","DOI":"10.1038\/s41586-025-09422-z"},{"key":"ref30","article-title":"Lmsys-chat-1m: A large-scale real-world llm conversation dataset","author":"Zheng","year":"2023"},{"key":"ref31","doi-asserted-by":"publisher","DOI":"10.1145\/3600006.3613165"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.52202\/068431-1189"},{"key":"ref33","article-title":"Sarathi: Efficient llm inference by piggybacking decodes with chunked prefills","author":"Agrawal","year":"2023"},{"key":"ref34","doi-asserted-by":"publisher","DOI":"10.1145\/3757892.3757909"},{"key":"ref35","article-title":"\" give me bf16 or give me death\u201d? accuracy-performance trade-offs in llm quantization","author":"Kurtic","year":"2024"},{"key":"ref36","article-title":"Mlperf power: Benchmarking the energy efficiency of machine learning systems from microwatts to megawatts for sustainable ai","author":"Tschand","year":"2024"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.1109\/CCGRID64434.2025.00035"},{"key":"ref38","first-page":"87","article-title":"Awq: Activation-aware weight quanti-zation for on-device llm compression and acceleration","volume-title":"Proceedings of machine learning and systems","volume":"6","author":"Lin"},{"key":"ref39","doi-asserted-by":"publisher","DOI":"10.52202\/075280-0441"},{"key":"ref40","first-page":"193","article-title":"{DistServe}: Disaggregating prefill and decoding for goodput-optimized large language model serving","volume-title":"18th USENIX Symposium on Operating Systems Design and Implementation (OSDI 24)","author":"Zhong"},{"key":"ref41","doi-asserted-by":"publisher","DOI":"10.52202\/079017-1996"},{"key":"ref42","doi-asserted-by":"publisher","DOI":"10.52202\/075280-1516"}],"event":{"name":"2026 IEEE International Parallel and Distributed Processing Symposium (IPDPS)","location":"New Orleans, LA, USA","start":{"date-parts":[[2026,5,25]]},"end":{"date-parts":[[2026,5,29]]}},"container-title":["2026 IEEE International Parallel and Distributed Processing Symposium (IPDPS)"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/11575315\/11575316\/11575329.pdf?arnumber=11575329","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,7,1]],"date-time":"2026-07-01T08:29:58Z","timestamp":1782894598000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11575329\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,5,25]]},"references-count":42,"URL":"https:\/\/doi.org\/10.1109\/ipdps65963.2026.00083","relation":{},"subject":[],"published":{"date-parts":[[2026,5,25]]}}}