{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,3,31]],"date-time":"2026-03-31T13:38:52Z","timestamp":1774964332279,"version":"3.50.1"},"reference-count":37,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","issue":"2","license":[{"start":{"date-parts":[[2025,6,1]],"date-time":"2025-06-01T00:00:00Z","timestamp":1748736000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2025,6,1]],"date-time":"2025-06-01T00:00:00Z","timestamp":1748736000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2025,6,1]],"date-time":"2025-06-01T00:00:00Z","timestamp":1748736000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"DOI":"10.13039\/100019186","name":"European Union’s Horizon Europe Research and Innovation Programme","doi-asserted-by":"publisher","award":["101070408"],"award-info":[{"award-number":["101070408"]}],"id":[{"id":"10.13039\/100019186","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100007569","name":"Carl-Zeiss Stiftung through the Sustainable Embedded AI Project","doi-asserted-by":"publisher","award":["P2021-02-009"],"award-info":[{"award-number":["P2021-02-009"]}],"id":[{"id":"10.13039\/100007569","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE Trans. Circuits Syst. Artif. Intel."],"published-print":{"date-parts":[[2025,6]]},"DOI":"10.1109\/tcasai.2025.3532254","type":"journal-article","created":{"date-parts":[[2025,2,10]],"date-time":"2025-02-10T13:29:31Z","timestamp":1739194171000},"page":"150-161","source":"Crossref","is-referenced-by-count":1,"title":["Efficient Deep Neural Network Training With a Novel 5.3-Bit Block Floating Point Data Format"],"prefix":"10.1109","volume":"2","author":[{"ORCID":"https:\/\/orcid.org\/0009-0009-6946-3330","authenticated-orcid":false,"given":"Mohammad Hassani","family":"Sadi","sequence":"first","affiliation":[{"name":"Microelectronic Systems Design Research Group, RPTU Kaiserslautern-Landau, Kaiserslautern, Germany"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-1651-1935","authenticated-orcid":false,"given":"Chirag","family":"Sudarshan","sequence":"additional","affiliation":[{"name":"Peter Grünberg Institute (PGI-14), Forschungszentrum Jülich GmbH, Jülich, Germany"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-5096-4794","authenticated-orcid":false,"given":"Sani R.","family":"Nassif","sequence":"additional","affiliation":[{"name":"Radyalis LLC, Austin, TX, USA"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-9010-086X","authenticated-orcid":false,"given":"Norbert","family":"Wehn","sequence":"additional","affiliation":[{"name":"Microelectronic Systems Design Research Group, RPTU Kaiserslautern-Landau, Kaiserslautern, Germany"}]}],"member":"263","reference":[{"key":"ref1","article-title":"Carbon emissions and large neural network training","author":"Patterson","year":"2021"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.1109\/MM.2024.3373763"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.1109\/ISCA52012.2021.00010"},{"key":"ref4","article-title":"NVIDIA a100 tensor core GPU","year":"2020"},{"key":"ref5","article-title":"NVIDIA h200 tensor core GPU","year":"2024"},{"key":"ref6","article-title":"Generative AI: The next s-curve for the semiconductor industry?"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1145\/3470496.3527408"},{"key":"ref8","article-title":"Product environmental report iPhone 11 Pro Max","year":"2019"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1145\/3620665.3640415"},{"key":"ref10","article-title":"Hybrid 8-bit floating point (HFP8) training and inference for deep neural networks","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"32","author":"Sun","year":"2019"},{"key":"ref11","article-title":"Rethinking floating point for deep learning","author":"Johnson","year":"2018"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.1109\/JSSC.2021.3103603"},{"key":"ref13","article-title":"FP8 formats for deep learning","author":"Micikevicius","year":"2022"},{"key":"ref14","article-title":"Training DNNs with hybrid block floating point","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"31","author":"Drumond","year":"2018"},{"key":"ref15","article-title":"A block minifloat representation for training deep neural networks","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Fox","year":"2020"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.1109\/TEC.1962.5219391"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1145\/3579371.3589351"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1109\/TC.2020.2985971"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.1007\/s10617-024-09282-2"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.1109\/EMC2-NIPS53020.2019.00010"},{"issue":"7","key":"ref21","first-page":"3","article-title":"Tiny imagenet visual recognition challenge","volume":"7","author":"Le","year":"2015","journal-title":"CS 231N"},{"key":"ref22","article-title":"Pytorch-cifar100","year":"2024"},{"key":"ref23","article-title":"Attention is all you need","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Vaswani","year":"2017"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.1109\/ISCAS48785.2022.9937832"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.1109\/HPCA51647.2021.00076"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1109\/MM.2022.3195617"},{"key":"ref27","doi-asserted-by":"publisher","DOI":"10.1016\/j.vlsi.2020.09.008"},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.1016\/j.eng.2020.01.007"},{"key":"ref29","article-title":"A study of BFLOAT16 for deep learning training","author":"Kalamkar","year":"2019"},{"key":"ref30","article-title":"Train with mixed precision documentation.\u201d"},{"key":"ref31","article-title":"Mixed precision training","author":"Micikevicius","year":"2017"},{"key":"ref32","article-title":"Training deep neural networks with 8-bit floating point numbers","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"31","author":"Wang","year":"2018"},{"key":"ref33","article-title":"Transformer engine documentation"},{"key":"ref34","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-15074-6_23"},{"key":"ref35","first-page":"10271","article-title":"Pushing the limits of narrow precision inferencing at cloud scale with microsoft floating point","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"33","author":"Rouhani","year":"2020"},{"key":"ref36","doi-asserted-by":"publisher","DOI":"10.1109\/TC.2023.3253050"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.1109\/HPCA53966.2022.00067"}],"container-title":["IEEE Transactions on Circuits and Systems for Artificial Intelligence"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/10495160\/11029988\/10878478.pdf?arnumber=10878478","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,12,9]],"date-time":"2025-12-09T18:32:13Z","timestamp":1765305133000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/10878478\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,6]]},"references-count":37,"journal-issue":{"issue":"2"},"URL":"https:\/\/doi.org\/10.1109\/tcasai.2025.3532254","relation":{},"ISSN":["2996-6647"],"issn-type":[{"value":"2996-6647","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,6]]}}}