{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,4,1]],"date-time":"2026-04-01T18:15:21Z","timestamp":1775067321824,"version":"3.50.1"},"reference-count":19,"publisher":"IEEE","license":[{"start":{"date-parts":[[2023,8,7]],"date-time":"2023-08-07T00:00:00Z","timestamp":1691366400000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2023,8,7]],"date-time":"2023-08-07T00:00:00Z","timestamp":1691366400000},"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":[],"published-print":{"date-parts":[[2023,8,7]]},"DOI":"10.1109\/islped58423.2023.10244348","type":"proceedings-article","created":{"date-parts":[[2023,9,19]],"date-time":"2023-09-19T17:38:53Z","timestamp":1695145133000},"page":"1-6","source":"Crossref","is-referenced-by-count":19,"title":["ITA: An Energy-Efficient Attention and Softmax Accelerator for Quantized Transformers"],"prefix":"10.1109","author":[{"given":"Gamze","family":"Islamoglu","sequence":"first","affiliation":[{"name":"ETH Zürich,Switzerland"}]},{"given":"Moritz","family":"Scherer","sequence":"additional","affiliation":[{"name":"ETH Zürich,Switzerland"}]},{"given":"Gianna","family":"Paulin","sequence":"additional","affiliation":[{"name":"ETH Zürich,Switzerland"}]},{"given":"Tim","family":"Fischer","sequence":"additional","affiliation":[{"name":"ETH Zürich,Switzerland"}]},{"given":"Victor J.B.","family":"Jung","sequence":"additional","affiliation":[{"name":"ETH Zürich,Switzerland"}]},{"given":"Angelo","family":"Garofalo","sequence":"additional","affiliation":[{"name":"ETH Zürich,Switzerland"}]},{"given":"Luca","family":"Benini","sequence":"additional","affiliation":[{"name":"ETH Zürich,Switzerland"}]}],"member":"263","reference":[{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1109\/VLSITechnologyandCir46769.2022.9830277"},{"key":"ref12","first-page":"1","article-title":"A 28nm 27.5TOPS\/W approximate-computing-based transformer processor with asymptotic sparsity specu-lating and out-of-order computing","volume":"65","author":"wang","year":"2022","journal-title":"2022 IEEE Int Solid-State Circuits Conf (ISSCC)"},{"key":"ref15","first-page":"97","author":"wang","year":"2021","journal-title":"SpAtten Efficient sparse attention architecture with cascade token and head pruning"},{"key":"ref14","article-title":"OPTIMUS: Op-timized matrix multiplication structure for transformer neural network accelerator","author":"park","year":"2020","journal-title":"MLSys"},{"key":"ref11","author":"bhandare","year":"2019","journal-title":"Efficient 8-bit quantization of transformer neural machine language translation model"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.1109\/EMC2-NIPS53020.2019.00016"},{"key":"ref2","author":"devlin","year":"2018","journal-title":"Bert Pretraining of deep bidirectional transformers for language understanding"},{"key":"ref1","author":"vaswani","year":"2017","journal-title":"Attention is all you need"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1109\/DAC18074.2021.9586134"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.1109\/ISCA52012.2021.00060"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.23919\/DATE51398.2021.9474087"},{"key":"ref18","author":"hassani","year":"2021","journal-title":"Escaping the big data paradigm with compact transformers"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1109\/COINS51742.2021.9524173"},{"key":"ref7","author":"benaich","year":"2022","journal-title":"State of AI report 2022"},{"key":"ref9","author":"kim","year":"2021","journal-title":"I-BERT integer-only BERT quantization"},{"key":"ref4","author":"dosovitskiy","year":"2020","journal-title":"An image is worth 16x16 words Transformers for image recognition at scale"},{"key":"ref3","author":"radford","year":"2019","journal-title":"Language Models are Unsupervised Multitask Learners"},{"key":"ref6","author":"bertasius","year":"2021","journal-title":"Is space-time attention all you need for video understanding?"},{"key":"ref5","author":"verma","year":"2021","journal-title":"Audio transformers transformer architectures for large scale audio understanding adieu convolutions"}],"event":{"name":"2023 IEEE\/ACM International Symposium on Low Power Electronics and Design (ISLPED)","location":"Vienna, Austria","start":{"date-parts":[[2023,8,7]]},"end":{"date-parts":[[2023,8,8]]}},"container-title":["2023 IEEE\/ACM International Symposium on Low Power Electronics and Design (ISLPED)"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx7\/10244176\/10244185\/10244348.pdf?arnumber=10244348","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2023,10,9]],"date-time":"2023-10-09T17:59:02Z","timestamp":1696874342000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/10244348\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2023,8,7]]},"references-count":19,"URL":"https:\/\/doi.org\/10.1109\/islped58423.2023.10244348","relation":{},"subject":[],"published":{"date-parts":[[2023,8,7]]}}}