{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,6,27]],"date-time":"2025-06-27T15:10:08Z","timestamp":1751037008773,"version":"3.41.0"},"publisher-location":"New York, NY, USA","reference-count":15,"publisher":"ACM","license":[{"start":{"date-parts":[[2026,6,29]],"date-time":"2026-06-29T00:00:00Z","timestamp":1782691200000},"content-version":"vor","delay-in-days":365,"URL":"http:\/\/www.acm.org\/publications\/policies\/copyright_policy#Background"}],"funder":[{"DOI":"10.13039\/100000015","name":"DOE U.S. Department of Energy","doi-asserted-by":"publisher","award":["DE-AC02-05CH11231"],"award-info":[{"award-number":["DE-AC02-05CH11231"]}],"id":[{"id":"10.13039\/100000015","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["dl.acm.org"],"crossmark-restriction":true},"short-container-title":[],"published-print":{"date-parts":[[2025,6,30]]},"DOI":"10.1145\/3716368.3735246","type":"proceedings-article","created":{"date-parts":[[2025,6,27]],"date-time":"2025-06-27T14:00:26Z","timestamp":1751032826000},"page":"718-725","update-policy":"https:\/\/doi.org\/10.1145\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["InFormer: A High-throughput, Ultra-efficient In-memory Compute-based Floating-point Arithmetic Accelerator for Transformers"],"prefix":"10.1145","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-0549-8341","authenticated-orcid":false,"given":"Hasita","family":"Veluri","sequence":"first","affiliation":[{"name":"Lawrence Berkeley National Laboratory, Berkeley, California, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-0931-309X","authenticated-orcid":false,"given":"Dilip","family":"Vasudevan","sequence":"additional","affiliation":[{"name":"Lawrence Berkeley National Laboratory, Berkeley, California, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"320","published-online":{"date-parts":[[2025,6,29]]},"reference":[{"key":"e_1_3_3_2_2_2","unstructured":"Ashish Vaswani Noam Shazeer Niki Parmar Jakob Uszkoreit Llion Jones Aidan\u00a0N. Gomez \u0141ukasz Kaiser and Illia Polosukhin. 2017. transformer: Attention is all you need. Advances in Neural Information Processing Systems 2017-Decem Nips (2017) 5999\u20136009."},{"key":"e_1_3_3_2_3_2","doi-asserted-by":"publisher","DOI":"10.1145\/3400302.3415640"},{"key":"e_1_3_3_2_4_2","doi-asserted-by":"publisher","unstructured":"Zhaojun Lu Xueyan Wang Md\u00a0Tanvir Arafin Haoxiang Yang Zhenglin Liu Jiliang Zhang and Gang Qu. 2024. An RRAM-Based Computing-in-Memory Architecture and Its Application in Accelerating Transformer Inference. IEEE Transactions on Very Large Scale Integration (VLSI) Systems 32 3 (2024) 485\u2013496. 10.1109\/TVLSI.2023.3345651","DOI":"10.1109\/TVLSI.2023.3345651"},{"key":"e_1_3_3_2_5_2","doi-asserted-by":"crossref","unstructured":"Nathan Leroux Paul-Philipp Manea Chirag Sudarshan Jan Finkbeiner Sebastian Siegel John\u00a0Paul Strachan and Emre Neftci. 2024. Analog In-Memory Computing Attention Mechanism for Fast and Energy-Efficient Large Language Models. (2024). http:\/\/arxiv.org\/abs\/2409.19315","DOI":"10.21203\/rs.3.rs-5461915\/v1"},{"key":"e_1_3_3_2_6_2","unstructured":"Christopher Wolters Xiaoxuan Yang Ulf Schlichtmann and Toyotaro Suzumura. 2024. Memory Is All You Need: An Overview of Compute-in-Memory Architectures for Accelerating Large Language Model Inference. (2024). http:\/\/arxiv.org\/abs\/2406.08413"},{"key":"e_1_3_3_2_7_2","doi-asserted-by":"publisher","unstructured":"Hasita Veluri Yida Li Jessie\u00a0Xuhua Niu Evgeny Zamburg and Aaron Voon\u00a0Yew Thean. 2021. High Throughput Area-Efficient and Variation-Tolerant 3D In-memory Compute System for Deep Convolutional Neural Networks. IEEE Internet of Things Journal 4662 c (2021) 1\u201315. 10.1109\/JIOT.2021.3058015","DOI":"10.1109\/JIOT.2021.3058015"},{"key":"e_1_3_3_2_8_2","doi-asserted-by":"publisher","DOI":"10.1109\/ISCA.2016.12"},{"key":"e_1_3_3_2_9_2","doi-asserted-by":"publisher","DOI":"10.1109\/MICRO56248.2022.00093"},{"key":"e_1_3_3_2_10_2","doi-asserted-by":"publisher","DOI":"10.1109\/isqed60706.2024.10528760"},{"key":"e_1_3_3_2_11_2","doi-asserted-by":"publisher","unstructured":"Miao Hu Hai Li Qing Wu and Garrett\u00a0S. Rose. 2012. Hardware realization of BSB recall function using memristor crossbar arrays. Proceedings - Design Automation Conference (2012) 498\u2013503. 10.1145\/2228360.2228448","DOI":"10.1145\/2228360.2228448"},{"key":"e_1_3_3_2_12_2","doi-asserted-by":"publisher","DOI":"10.1145\/3307650.3322237"},{"key":"e_1_3_3_2_13_2","doi-asserted-by":"publisher","unstructured":"Hasita Veluri Umesh Chand Yida Li Baoshan Tang and Aaron Voon\u00a0Yew Thean. 2023. A Low-Power DNN Accelerator Enabled by a Novel Staircase RRAM Array. IEEE Transactions on Neural Networks and Learning Systems 34 8 (2023) 4416\u20134427. 10.1109\/TNNLS.2021.3118451","DOI":"10.1109\/TNNLS.2021.3118451"},{"key":"e_1_3_3_2_14_2","doi-asserted-by":"publisher","DOI":"10.1145\/3410463.3414634"},{"key":"e_1_3_3_2_15_2","doi-asserted-by":"publisher","unstructured":"Peng Lin Can Li Zhongrui Wang Yunning Li Hao Jiang Wenhao Song Mingyi Rao Ye Zhuo Navnidhi\u00a0K. Upadhyay Mark Barnell Qing Wu J.\u00a0Joshua Yang and Qiangfei Xia. 2020. Three-dimensional memristor circuits as complex neural networks. Nature Electronics 3 4 (2020) 225\u2013232. 10.1038\/s41928-020-0397-9","DOI":"10.1038\/s41928-020-0397-9"},{"key":"e_1_3_3_2_16_2","doi-asserted-by":"publisher","DOI":"10.1109\/cnna.2010.5430304"}],"event":{"name":"GLSVLSI '25: Great Lakes Symposium on VLSI 2025","sponsor":["SIGDA ACM Special Interest Group on Design Automation"],"location":"New Orleans LA USA","acronym":"GLSVLSI '25"},"container-title":["Proceedings of the Great Lakes Symposium on VLSI 2025"],"original-title":[],"link":[{"URL":"https:\/\/dl.acm.org\/doi\/abs\/10.1145\/3716368.3735246","content-type":"text\/html","content-version":"vor","intended-application":"syndication"}],"deposited":{"date-parts":[[2025,6,27]],"date-time":"2025-06-27T14:38:27Z","timestamp":1751035107000},"score":1,"resource":{"primary":{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3716368.3735246"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,6,29]]},"references-count":15,"alternative-id":["10.1145\/3716368.3735246","10.1145\/3716368"],"URL":"https:\/\/doi.org\/10.1145\/3716368.3735246","relation":{},"subject":[],"published":{"date-parts":[[2025,6,29]]},"assertion":[{"value":"2025-06-29","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}