{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,9,17]],"date-time":"2025-09-17T06:09:10Z","timestamp":1758089350757,"version":"3.44.0"},"reference-count":22,"publisher":"IEEE","license":[{"start":{"date-parts":[[2025,6,22]],"date-time":"2025-06-22T00:00:00Z","timestamp":1750550400000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2025,6,22]],"date-time":"2025-06-22T00:00:00Z","timestamp":1750550400000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"DOI":"10.13039\/100006190","name":"Research and Development","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100006190","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2025,6,22]]},"DOI":"10.1109\/dac63849.2025.11132475","type":"proceedings-article","created":{"date-parts":[[2025,9,15]],"date-time":"2025-09-15T17:35:41Z","timestamp":1757957741000},"page":"1-7","source":"Crossref","is-referenced-by-count":0,"title":["Blaze: An Efficient Bit-Sparse Attention Architecture With Workload Orchestration Optimization"],"prefix":"10.1109","author":[{"given":"Runzhou","family":"Zhang","sequence":"first","affiliation":[{"name":"Fudan University,State Key Lab of Integrated Chips & Systems, and School of Microelectronics,Shanghai,China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Faxian","family":"Sun","sequence":"additional","affiliation":[{"name":"Fudan University,State Key Lab of Integrated Chips & Systems, and School of Microelectronics,Shanghai,China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Yiming","family":"Wang","sequence":"additional","affiliation":[{"name":"Fudan University,State Key Lab of Integrated Chips & Systems, and School of Microelectronics,Shanghai,China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Kunchen","family":"Zou","sequence":"additional","affiliation":[{"name":"Fudan University,State Key Lab of Integrated Chips & Systems, and School of Microelectronics,Shanghai,China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Zhinan","family":"Qin","sequence":"additional","affiliation":[{"name":"Fudan University,State Key Lab of Integrated Chips & Systems, and School of Microelectronics,Shanghai,China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Jianli","family":"Chen","sequence":"additional","affiliation":[{"name":"Fudan University,State Key Lab of Integrated Chips & Systems, and School of Microelectronics,Shanghai,China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Jun","family":"Yu","sequence":"additional","affiliation":[{"name":"Fudan University,State Key Lab of Integrated Chips & Systems, and School of Microelectronics,Shanghai,China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Kun","family":"Wang","sequence":"additional","affiliation":[{"name":"Fudan University,State Key Lab of Integrated Chips & Systems, and School of Microelectronics,Shanghai,China"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"263","reference":[{"key":"ref1","article-title":"Attention is all you need","author":"Vaswani","year":"2017","journal-title":"Advances in Neural Information Processing Systems"},{"key":"ref2","article-title":"Bert: Pre-training of deep bidirectional transformers for language understanding","author":"Devlin","year":"2018","journal-title":"arXiv preprint arXiv:1810.04805"},{"issue":"8","key":"ref3","first-page":"9","article-title":"Language models are unsupervised multitask learners","volume":"1","author":"Radford","year":"2019","journal-title":"OpenAI blog"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.1109\/HPCA51647.2021.00018"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1145\/3466752.3480125"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.23919\/DATE54114.2022.9774692"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1109\/TCAD.2022.3181541"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1109\/ICCD58817.2023.00098"},{"key":"ref9","first-page":"382","article-title":"Bit-pragmatic deep neural network computing","volume-title":"Proceedings of the 50th annual IEEE\/ACM international symposium on microarchitecture","author":"Albericio"},{"key":"ref10","first-page":"304","article-title":"Laconic deep learning inference acceleration","volume-title":"Proceedings of the 46th International Symposium on Computer Architecture","author":"Sharify"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1109\/DAC56929.2023.10247662"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.1109\/tc.2023.3324477"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1109\/TCAD.2022.3146202"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.1109\/TCSI.2024.3426653"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2010.11929"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.1145\/3352460.3358291"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/D16-1264"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/d18-1257"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/W18-5446"},{"key":"ref20","article-title":"Pytorch: An imperative style, high-performance deep learning library","volume":"32","author":"Paszke","year":"2019","journal-title":"Advances in neural information processing systems"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2020.emnlp-demos.6"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.1145\/3085572"}],"event":{"name":"2025 62nd ACM\/IEEE Design Automation Conference (DAC)","start":{"date-parts":[[2025,6,22]]},"location":"San Francisco, CA, USA","end":{"date-parts":[[2025,6,25]]}},"container-title":["2025 62nd ACM\/IEEE Design Automation Conference (DAC)"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/11132383\/11132091\/11132475.pdf?arnumber=11132475","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,9,16]],"date-time":"2025-09-16T05:38:19Z","timestamp":1758001099000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11132475\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,6,22]]},"references-count":22,"URL":"https:\/\/doi.org\/10.1109\/dac63849.2025.11132475","relation":{},"subject":[],"published":{"date-parts":[[2025,6,22]]}}}