{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,9]],"date-time":"2026-05-09T07:05:30Z","timestamp":1778310330473,"version":"3.51.4"},"reference-count":10,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","issue":"6","license":[{"start":{"date-parts":[[2024,11,1]],"date-time":"2024-11-01T00:00:00Z","timestamp":1730419200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2024,11,1]],"date-time":"2024-11-01T00:00:00Z","timestamp":1730419200000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2024,11,1]],"date-time":"2024-11-01T00:00:00Z","timestamp":1730419200000},"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":["IEEE Micro"],"published-print":{"date-parts":[[2024,11]]},"DOI":"10.1109\/mm.2024.3420728","type":"journal-article","created":{"date-parts":[[2024,7,9]],"date-time":"2024-07-09T18:29:46Z","timestamp":1720549786000},"page":"17-33","source":"Crossref","is-referenced-by-count":11,"title":["A Latency Processing Unit: A Latency-Optimized and Highly Scalable Processor for Large Language Model Inference"],"prefix":"10.1109","volume":"44","author":[{"ORCID":"https:\/\/orcid.org\/0009-0002-5924-7000","authenticated-orcid":false,"given":"Seungjae","family":"Moon","sequence":"first","affiliation":[{"name":"HyperAccel, Seoul, South Korea"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-4702-6489","authenticated-orcid":false,"given":"Jung-Hoon","family":"Kim","sequence":"additional","affiliation":[{"name":"Korea Advanced Institute of Science Technology, Daejeon, South Korea"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-6680-2602","authenticated-orcid":false,"given":"Junsoo","family":"Kim","sequence":"additional","affiliation":[{"name":"HyperAccel, Seoul, South Korea"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0005-7940-8221","authenticated-orcid":false,"given":"Seongmin","family":"Hong","sequence":"additional","affiliation":[{"name":"HyperAccel, Seoul, South Korea"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0004-4415-8552","authenticated-orcid":false,"given":"Junseo","family":"Cha","sequence":"additional","affiliation":[{"name":"HyperAccel, Seoul, South Korea"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0003-8751-0352","authenticated-orcid":false,"given":"Minsu","family":"Kim","sequence":"additional","affiliation":[{"name":"Korea Advanced Institute of Science Technology, Daejeon, South Korea"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-5743-4852","authenticated-orcid":false,"given":"Sukbin","family":"Lim","sequence":"additional","affiliation":[{"name":"HyperAccel, Seoul, South Korea"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0008-0759-5357","authenticated-orcid":false,"given":"Gyubin","family":"Choi","sequence":"additional","affiliation":[{"name":"HyperAccel, Seoul, South Korea"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0002-3404-9542","authenticated-orcid":false,"given":"Dongjin","family":"Seo","sequence":"additional","affiliation":[{"name":"HyperAccel, Seoul, South Korea"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0009-8219-1317","authenticated-orcid":false,"given":"Jongho","family":"Kim","sequence":"additional","affiliation":[{"name":"HyperAccel, Seoul, South Korea"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0006-4460-9530","authenticated-orcid":false,"given":"Hunjong","family":"Lee","sequence":"additional","affiliation":[{"name":"HyperAccel, Seoul, South Korea"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0001-4036-7590","authenticated-orcid":false,"given":"Hyunjun","family":"Park","sequence":"additional","affiliation":[{"name":"HyperAccel, Seoul, South Korea"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0006-3829-7761","authenticated-orcid":false,"given":"Ryeowook","family":"Ko","sequence":"additional","affiliation":[{"name":"HyperAccel, Seoul, South Korea"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0005-6453-9755","authenticated-orcid":false,"given":"Soongyu","family":"Choi","sequence":"additional","affiliation":[{"name":"HyperAccel, Seoul, South Korea"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-6629-449X","authenticated-orcid":false,"given":"Jongse","family":"Park","sequence":"additional","affiliation":[{"name":"Korea Advanced Institute of Science Technology, Daejeon, South Korea"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0003-1830-634X","authenticated-orcid":false,"given":"Jinwon","family":"Lee","sequence":"additional","affiliation":[{"name":"HyperAccel, Seoul, South Korea"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-1099-1496","authenticated-orcid":false,"given":"Joo-Young","family":"Kim","sequence":"additional","affiliation":[{"name":"HyperAccel, Seoul, South Korea"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.48550\/ARXIV.1706.03762"},{"issue":"1","key":"ref2","first-page":"11,324","article-title":"Palm: Scaling language modeling with pathways","volume":"24","author":"Chowdhery","year":"2023","journal-title":"J. Mach. Learn. Res. (JMLR)"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.1109\/hpca51647.2021.00018"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2020.emnlp-demos.6"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1109\/micro56248.2022.00051"},{"key":"ref6","first-page":"3974","article-title":"On-the-fly operation batching in dynamic computation graphs","volume-title":"Proc. 31st Conf. Neural Inf. Process. Syst. (NIPS)","author":"Neubig","year":"2017"},{"key":"ref7","article-title":"Reformer: The efficient transformer","author":"Kitaev","year":"2020"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1109\/LCA.2015.2414456"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1109\/mm.2020.2975185"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.1145\/3470496.3527405"}],"container-title":["IEEE Micro"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/40\/10812075\/10591630.pdf?arnumber=10591630","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,12,30]],"date-time":"2024-12-30T19:28:20Z","timestamp":1735586900000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/10591630\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,11]]},"references-count":10,"journal-issue":{"issue":"6"},"URL":"https:\/\/doi.org\/10.1109\/mm.2024.3420728","relation":{},"ISSN":["0272-1732","1937-4143"],"issn-type":[{"value":"0272-1732","type":"print"},{"value":"1937-4143","type":"electronic"}],"subject":[],"published":{"date-parts":[[2024,11]]}}}