{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,4,13]],"date-time":"2026-04-13T23:14:41Z","timestamp":1776122081190,"version":"3.50.1"},"reference-count":31,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","issue":"7","license":[{"start":{"date-parts":[[2023,7,1]],"date-time":"2023-07-01T00:00:00Z","timestamp":1688169600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2023,7,1]],"date-time":"2023-07-01T00:00:00Z","timestamp":1688169600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2023,7,1]],"date-time":"2023-07-01T00:00:00Z","timestamp":1688169600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"name":"National Key R&D Program of China","award":["2021YFB0301300"],"award-info":[{"award-number":["2021YFB0301300"]}]},{"name":"Zhejiang Lab","award":["2021KC0AB04"],"award-info":[{"award-number":["2021KC0AB04"]}]},{"name":"Major Program of Guangdong Basic and Applied Research","award":["2019B030302002"],"award-info":[{"award-number":["2019B030302002"]}]},{"name":"Program for Guangdong Natural Science Foundation","award":["2018B030312002"],"award-info":[{"award-number":["2018B030312002"]}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE Trans. Parallel Distrib. Syst."],"published-print":{"date-parts":[[2023,7]]},"DOI":"10.1109\/tpds.2023.3280805","type":"journal-article","created":{"date-parts":[[2023,5,29]],"date-time":"2023-05-29T17:44:04Z","timestamp":1685382244000},"page":"2221-2235","source":"Crossref","is-referenced-by-count":16,"title":["Full-Stack Optimizing Transformer Inference on ARM Many-Core CPU"],"prefix":"10.1109","volume":"34","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-1417-3012","authenticated-orcid":false,"given":"Jiazhi","family":"Jiang","sequence":"first","affiliation":[{"name":"School of Computer Science and Engineering, Sun Yat-sen University, Guangzhou, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-4707-9492","authenticated-orcid":false,"given":"Jiangsu","family":"Du","sequence":"additional","affiliation":[{"name":"School of Computer Science and Engineering, Sun Yat-sen University, Guangzhou, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-5582-1031","authenticated-orcid":false,"given":"Dan","family":"Huang","sequence":"additional","affiliation":[{"name":"School of Computer Science and Engineering, Sun Yat-sen University, Guangzhou, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-9318-5715","authenticated-orcid":false,"given":"Zhiguang","family":"Chen","sequence":"additional","affiliation":[{"name":"School of Computer Science and Engineering, Sun Yat-sen University, Guangzhou, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Yutong","family":"Lu","sequence":"additional","affiliation":[{"name":"School of Computer Science and Engineering, Sun Yat-sen University, Guangzhou, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Xiangke","family":"Liao","sequence":"additional","affiliation":[{"name":"School of Computer Science and Engineering, Sun Yat-sen University, Guangzhou, China"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"263","reference":[{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1145\/3330345.3330355"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.1145\/3458817.3476217"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.1145\/3232162"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.1145\/3545008.3545022"},{"key":"ref31","article-title":"RoBERTa: A robustly optimized bert pretraining approach","author":"liu","year":"2019"},{"key":"ref30","article-title":"Distilbert, a distilled version of BERT: Smaller, faster, cheaper and lighter","author":"sanh","year":"2019"},{"key":"ref11","first-page":"8026","article-title":"PyTorch: An imperative style, high-performance deep learning library","author":"paszke","year":"2019","journal-title":"Proc Adv Neural Inf Process Syst"},{"key":"ref10","year":"2021"},{"key":"ref2","article-title":"Language models are unsupervised multitask learners","volume":"1","author":"radford","year":"2019","journal-title":"OpenAIRE blog"},{"key":"ref1","article-title":"BERT: Pre-training of deep bidirectional transformers for language understanding","author":"devlin","year":"2018"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.3390\/electronics10161984"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.1016\/j.future.2020.06.033"},{"key":"ref19","first-page":"1","article-title":"ET: Re-thinking self-attention for transformer models on GPUs","author":"chen","year":"2021","journal-title":"Proc Int Conf High Perform Comput Netw Storage Anal"},{"key":"ref18","year":"2021"},{"key":"ref24","article-title":"Optimizing inference performance of transformers on CPUs","author":"dice","year":"2021"},{"key":"ref23","article-title":"MXNet: A flexible and efficient machine learning library for heterogeneous distributed systems","author":"chen","year":"2015"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1109\/TPDS.2020.3030548"},{"key":"ref25","article-title":"Microsoft open sources breakthrough optimizations for transformer inference on GPU and CPU","author":"ning","year":"0"},{"key":"ref20","first-page":"36","article-title":"Q8BERT: Quantized 8 bit BERT","author":"zafrir","year":"2019","journal-title":"Proc IEEE 5th Workshop Energy Efficient Mach Learn Cogn Comput"},{"key":"ref22","article-title":"Optimization for bert inference performance on CPU","author":"wu","year":"0"},{"key":"ref21","article-title":"Flashattention: Fast and memory-efficient exact attention with IO-awareness","author":"dao","year":"2022"},{"key":"ref28","first-page":"3393","article-title":"Learning to optimize tensor programs","author":"chen","year":"2018","journal-title":"Proc Adv Neural Inf Process Syst"},{"key":"ref27","article-title":"TVM: An automated end-to-end optimizing compiler for deep learning","author":"chen","year":"2018"},{"key":"ref29","doi-asserted-by":"publisher","DOI":"10.1145\/1377603.1377607"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1145\/3472456.3472496"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1109\/HPEC43674.2020.9286150"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1145\/3437801.3441578"},{"key":"ref4","article-title":"Exploring the role of machine learning in scientific workflows: Opportunities and challenges","author":"nouri","year":"2021"},{"key":"ref3","first-page":"495","article-title":"Dorylus: Affordable, scalable, and accurate $\\lbrace${GNN$\\rbrace$} training with distributed $\\lbrace${CPU$\\rbrace$} servers and serverless threads","author":"thorpe","year":"0","journal-title":"Proc 15th USENIX Symp Operating Syst Des Implementation"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.1109\/SC41405.2020.00051"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1007\/s11432-016-5588-7"}],"container-title":["IEEE Transactions on Parallel and Distributed Systems"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx7\/71\/10132303\/10138027.pdf?arnumber=10138027","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2023,6,26]],"date-time":"2023-06-26T18:51:18Z","timestamp":1687805478000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/10138027\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2023,7]]},"references-count":31,"journal-issue":{"issue":"7"},"URL":"https:\/\/doi.org\/10.1109\/tpds.2023.3280805","relation":{},"ISSN":["1045-9219","1558-2183","2161-9883"],"issn-type":[{"value":"1045-9219","type":"print"},{"value":"1558-2183","type":"electronic"},{"value":"2161-9883","type":"electronic"}],"subject":[],"published":{"date-parts":[[2023,7]]}}}