{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,1]],"date-time":"2026-07-01T12:49:14Z","timestamp":1782910154183,"version":"3.54.5"},"reference-count":31,"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\/100000015","name":"U.S. Department of Energy","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100000015","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100006132","name":"Office of Science","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100006132","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100006192","name":"Advanced Scientific Computing Research","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100006192","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100000001","name":"National Science Foundation","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100000001","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\/ipdps65963.2026.00057","type":"proceedings-article","created":{"date-parts":[[2026,6,30]],"date-time":"2026-06-30T20:55:16Z","timestamp":1782852916000},"page":"598-611","source":"Crossref","is-referenced-by-count":0,"title":["PackKV: Reducing KV Cache Memory Footprint through LLM-Aware Lossy Compression"],"prefix":"10.1109","author":[{"given":"Bo","family":"Jiang","sequence":"first","affiliation":[{"name":"Temple University,Philadelphia,USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Taolue","family":"Yang","sequence":"additional","affiliation":[{"name":"Temple University,Philadelphia,USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Youyuan","family":"Liu","sequence":"additional","affiliation":[{"name":"Temple University,Philadelphia,USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xubin","family":"He","sequence":"additional","affiliation":[{"name":"Temple University,Philadelphia,USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sheng","family":"Di","sequence":"additional","affiliation":[{"name":"Argonne National Laboratory,USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sian","family":"Jin","sequence":"additional","affiliation":[{"name":"Temple University,Philadelphia,USA"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","first-page":"1877","article-title":"Language models are few-shot learners","volume":"33","author":"Brown","year":"2020","journal-title":"Advances in neural information processing systems"},{"key":"ref2","article-title":"Galactica: A large language model for science","author":"Taylor","year":"2022"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.1145\/3600006.3613165"},{"key":"ref4","first-page":"606","article-title":"Efficiently scaling transformer inference","volume-title":"Proceedings of Machine Learning and Systems","volume":"5","author":"Pope"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.52202\/075280-2279"},{"key":"ref6","article-title":"Kivi: A tuning-free asymmetric 2bit quantization for kv cache","author":"Liu","year":"2024"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.52202\/079017-0040"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1109\/HPCA51647.2021.00018"},{"key":"ref9","first-page":"381","article-title":"Q-hitter: A better token oracle for efficient llm inference via sparse-quantized kv cache","volume-title":"Proceedings of Machine Learning and Systems","volume":"6","author":"Zhang"},{"key":"ref10","first-page":"31 094","article-title":"Flexgen: High-throughput generative inference of large language models with a single gpu","volume-title":"International Conference on Machine Learning","author":"Sheng"},{"key":"ref11","article-title":"Gpt-4 technical report","author":"Achiam","year":"2023"},{"key":"ref12","article-title":"Improving language understanding by generative pre-training","author":"Radford","year":"2018"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.48550\/ARXIV.1706.03762"},{"key":"ref14","article-title":"Fast transformer decoding: One write-head is all you need","author":"Shazeer","year":"2019"},{"key":"ref15","article-title":"Sarathi: Efficient llm inference by piggybacking decodes with chunked prefills","author":"Agrawal","year":"2023"},{"key":"ref16","article-title":"Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Coding","volume-title":"4th International Conference on Learning Representations (ICLR)","author":"Han"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2018.00286"},{"key":"ref18","article-title":"Post training 4-bit quantization of convolutional networks for rapid-deployment","volume":"32","author":"Banner","year":"2019","journal-title":"Advances in Neural Information Processing Systems"},{"key":"ref19","article-title":"Gptq: Accurate post-training quantization for generative pre-trained transformers","author":"Frantar","year":"2022"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.1145\/3307681.3326608"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.14778\/3503585.3503597"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.1109\/30.125072"},{"key":"ref23","doi-asserted-by":"publisher","DOI":"10.1016\/s0923-5965(01)00028-5"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.1109\/TVCG.2014.2346458"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.1109\/TCSVT.2012.2221192"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1109\/IPDPS.2016.11"},{"key":"ref27","doi-asserted-by":"publisher","DOI":"10.1109\/IPDPS.2017.115"},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.1109\/BigData.2018.8622520"},{"key":"ref29","doi-asserted-by":"publisher","DOI":"10.1145\/3651890.3672274"},{"key":"ref30","doi-asserted-by":"publisher","DOI":"10.1109\/TPDS.2016.2549523"},{"key":"ref31","article-title":"Optimisation-based solution methods for set partitioning models","author":"Rasmussen","year":"2011"}],"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\/11575427.pdf?arnumber=11575427","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,7,1]],"date-time":"2026-07-01T12:16:11Z","timestamp":1782908171000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11575427\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,5,25]]},"references-count":31,"URL":"https:\/\/doi.org\/10.1109\/ipdps65963.2026.00057","relation":{},"subject":[],"published":{"date-parts":[[2026,5,25]]}}}