{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,12]],"date-time":"2026-08-12T19:31:46Z","timestamp":1786563106262,"version":"build-2736575974"},"reference-count":52,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","issue":"9","license":[{"start":{"date-parts":[[2026,9,1]],"date-time":"2026-09-01T00:00:00Z","timestamp":1788220800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2026,9,1]],"date-time":"2026-09-01T00:00:00Z","timestamp":1788220800000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2026,9,1]],"date-time":"2026-09-01T00:00:00Z","timestamp":1788220800000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"crossref","award":["U22A2028"],"award-info":[{"award-number":["U22A2028"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"crossref"}]},{"name":"Strategic Priority Research Program of the Chinese Academy of Sciences","award":["XDB0660300"],"award-info":[{"award-number":["XDB0660300"]}]},{"name":"CAS Project for Young Scientists in Basic Research","award":["YSBR-029"],"award-info":[{"award-number":["YSBR-029"]}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE Trans. Comput."],"published-print":{"date-parts":[[2026,9]]},"DOI":"10.1109\/tc.2026.3710209","type":"journal-article","created":{"date-parts":[[2026,7,8]],"date-time":"2026-07-08T19:42:36Z","timestamp":1783539756000},"page":"3369-3382","source":"Crossref","is-referenced-by-count":0,"title":["Cambricon-QM: A Hybrid Architecture for Microscaling Format Training"],"prefix":"10.1109","volume":"75","author":[{"ORCID":"https:\/\/orcid.org\/0009-0005-1457-3672","authenticated-orcid":false,"given":"Ximing","family":"Liu","sequence":"first","affiliation":[{"name":"Institute of Computing Technology, Chinese Academy of Sciences","place":["Beijing, China"],"department":["State Key Lab of Processors"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-5503-4457","authenticated-orcid":false,"given":"Yongwei","family":"Zhao","sequence":"additional","affiliation":[{"name":"Institute of Computing Technology, Chinese Academy of Sciences","place":["Beijing, China"],"department":["State Key Lab of Processors"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-9540-6086","authenticated-orcid":false,"given":"Chang","family":"Liu","sequence":"additional","affiliation":[{"name":"Cambricon Technologies","place":["Beijing, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-7603-4210","authenticated-orcid":false,"given":"Zidong","family":"Du","sequence":"additional","affiliation":[{"name":"Institute of Computing Technology, Chinese Academy of Sciences","place":["Beijing, China"],"department":["State Key Lab of Processors"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-9979-0561","authenticated-orcid":false,"given":"Xing","family":"Hu","sequence":"additional","affiliation":[{"name":"Institute of Computing Technology, Chinese Academy of Sciences","place":["Beijing, China"],"department":["State Key Lab of Processors"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0007-8000-005X","authenticated-orcid":false,"given":"Yimin","family":"Zhuang","sequence":"additional","affiliation":[{"name":"Institute of Computing Technology, Chinese Academy of Sciences","place":["Beijing, China"],"department":["State Key Lab of Processors"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-3355-8361","authenticated-orcid":false,"given":"Zhenxing","family":"Zhang","sequence":"additional","affiliation":[{"name":"Institute of Computing Technology, Chinese Academy of Sciences","place":["Beijing, China"],"department":["State Key Lab of Processors"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0002-9823-2573","authenticated-orcid":false,"given":"Yifan","family":"Hao","sequence":"additional","affiliation":[{"name":"Institute of Computing Technology, Chinese Academy of Sciences","place":["Beijing, China"],"department":["State Key Lab of Processors"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-3458-1555","authenticated-orcid":false,"given":"Xinkai","family":"Song","sequence":"additional","affiliation":[{"name":"Institute of Computing Technology, Chinese Academy of Sciences","place":["Beijing, China"],"department":["State Key Lab of Processors"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-8691-8549","authenticated-orcid":false,"given":"Rui","family":"Zhang","sequence":"additional","affiliation":[{"name":"Institute of Computing Technology, Chinese Academy of Sciences","place":["Beijing, China"],"department":["State Key Lab of Processors"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0006-0237-1034","authenticated-orcid":false,"given":"Wei","family":"Li","sequence":"additional","affiliation":[{"name":"Institute of Computing Technology, Chinese Academy of Sciences","place":["Beijing, China"],"department":["State Key Lab of Processors"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0007-7607-9041","authenticated-orcid":false,"given":"Xishan","family":"Zhang","sequence":"additional","affiliation":[{"name":"Institute of Computing Technology, Chinese Academy of Sciences","place":["Beijing, China"],"department":["State Key Lab of Processors"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-8877-9052","authenticated-orcid":false,"given":"Ling","family":"Li","sequence":"additional","affiliation":[{"name":"University of Chinese Academy of Sciences","place":["Beijing, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-1480-7265","authenticated-orcid":false,"given":"Zhiwei","family":"Xu","sequence":"additional","affiliation":[{"name":"Institute of Computing Technology, Chinese Academy of Sciences","place":["Beijing, China"],"department":["State Key Lab of Processors"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-7601-0753","authenticated-orcid":false,"given":"Tianshi","family":"Chen","sequence":"additional","affiliation":[{"name":"Cambricon Technologies","place":["Beijing, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-2530-5874","authenticated-orcid":false,"given":"Qi","family":"Guo","sequence":"additional","affiliation":[{"name":"Institute of Computing Technology, Chinese Academy of Sciences","place":["Beijing, China"],"department":["State Key Lab of Processors"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","article-title":"Language models are few-shot learners","author":"Brown","year":"2020"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.1109\/MM.2023.3256796"},{"key":"ref3","article-title":"Mixed precision training","author":"Micikevicius","year":"2017"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.1145\/3079856.3080246"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1109\/ISSCC.2014.6757323"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR42600.2020.00240"},{"key":"ref7","first-page":"802","article-title":"FloatPIM: In-memory acceleration of deep neural network training with high precision","volume-title":"Proc. 46th Int. Symp. Comput. Archit.","author":"Imani","year":"2019"},{"key":"ref8","article-title":"Microscaling data formats for deep learning","author":"Rouhani","year":"2023"},{"issue":"7","key":"ref9","first-page":"2121","article-title":"Adaptive subgradient methods for online learning and stochastic optimization","volume":"12","author":"Duchi","year":"2011","journal-title":"J. Mach. Learn. Res."},{"key":"ref10","article-title":"Lecture 6e: Rmsprop: Divide the gradient by a running average of its recent magnitude","author":"Hinton","year":"2012"},{"key":"ref11","article-title":"Adam: A method for stochastic optimization","author":"Diederik","year":"2014"},{"key":"ref12","article-title":"Nvidia tesla v100 GPU architecture","year":"2018"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1109\/HPCA47549.2020.00015"},{"key":"ref14","doi-asserted-by":"crossref","first-page":"70","DOI":"10.1016\/j.neunet.2019.12.027","article-title":"Training high-performance and large-scale deep neural networks with full 8-bit integers","volume":"125","author":"Yang","year":"2020","journal-title":"Neural Netw."},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR42600.2020.00204"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.1109\/ISCA52012.2021.00061"},{"key":"ref17","article-title":"Llama: Open and efficient foundation language models","author":"Touvron","year":"2023"},{"key":"ref18","article-title":"Improving language understanding by generative pre-training","author":"Radford","year":"2018"},{"key":"ref19","article-title":"The Llama 3 herd of models","author":"Dubey","year":"2024"},{"key":"ref20","article-title":"BERT: Pre-training of deep bidirectional transformers for language understanding","author":"Devlin","year":"2018"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2016.90"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.1145\/3065386"},{"key":"ref23","doi-asserted-by":"publisher","DOI":"10.1145\/3085572"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.3390\/jlpea7030023"},{"key":"ref25","article-title":"Jetson AGX Xavier and the new era of autonomous machines","year":"2018"},{"key":"ref26","article-title":"Microxcaling"},{"key":"ref27","article-title":"Scale-SIM: Systolic CNN accelerator simulator","author":"Samajdar","year":"2018"},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.1145\/1498765.1498785"},{"key":"ref29","doi-asserted-by":"publisher","DOI":"10.1109\/IPDPSW.2018.00091"},{"key":"ref30","doi-asserted-by":"publisher","DOI":"10.1109\/MM.2021.3061394"},{"key":"ref31","doi-asserted-by":"publisher","DOI":"10.1145\/3297858.3304014"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.1109\/MICRO.2016.7783722"},{"key":"ref33","doi-asserted-by":"publisher","DOI":"10.1109\/ISCA.2018.00069"},{"key":"ref34","doi-asserted-by":"publisher","DOI":"10.1145\/2654822.2541967"},{"key":"ref35","doi-asserted-by":"publisher","DOI":"10.1145\/3007787.3001177"},{"issue":"187","key":"ref36","first-page":"1","article-title":"Quantized neural networks: Training neural networks with low precision weights and activations","volume":"18","author":"Hubara","year":"2018","journal-title":"J. Mach. Learn. Res."},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.23919\/DATE.2018.8342119"},{"key":"ref38","first-page":"451","article-title":"Training DNNs with hybrid block floating point","volume":"31","author":"Drumond","year":"2018","journal-title":"Proc. Adv. Neural Inf. Process. Syst."},{"key":"ref39","first-page":"10271","article-title":"Pushing the limits of narrow precision inferencing at cloud scale with Microsoft floating point","volume":"33","author":"Darvish Rouhani","year":"2020","journal-title":"Advances in Neural Inf. Process. Syst."},{"key":"ref40","article-title":"Ocp microscaling formats (mx) v1.0 specification","year":"2023"},{"key":"ref41","doi-asserted-by":"publisher","DOI":"10.1145\/3352460.3358295"},{"key":"ref42","doi-asserted-by":"publisher","DOI":"10.1109\/ISCA.2018.00063"},{"key":"ref43","doi-asserted-by":"publisher","DOI":"10.1109\/FPL.2018.00059"},{"key":"ref44","doi-asserted-by":"publisher","DOI":"10.1145\/3173162.3173177"},{"key":"ref45","doi-asserted-by":"publisher","DOI":"10.1109\/ISCA45697.2020.00086"},{"key":"ref46","first-page":"1","article-title":"Biscaled-DNN: Quantizing long-tailed datastructures with two scale factors for deep neural networks","volume-title":"Proc. 56th Annu. Des. Automat. Conf.","author":"Jain","year":"2019"},{"key":"ref47","article-title":"Mixed precision training of convolutional neural networks using integer operations","author":"Das","year":"2018"},{"key":"ref48","first-page":"1737","article-title":"Deep learning with limited numerical precision","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Gupta","year":"2015"},{"key":"ref49","doi-asserted-by":"publisher","DOI":"10.1109\/VLSIC.2018.8502276"},{"key":"ref50","doi-asserted-by":"publisher","DOI":"10.1109\/vlsicircuits18222.2020.9162917"},{"key":"ref51","doi-asserted-by":"publisher","DOI":"10.1109\/ISSCC.2019.8662302"},{"key":"ref52","first-page":"1002","article-title":"Bucket getter: A bucket-based processing engine for low-bit block floating point (BFP) DNNs","volume-title":"Proc. 56th Annu. IEEE\/ACM Int. Symp. Microarchit.","author":"Lo","year":"2023"}],"container-title":["IEEE Transactions on Computers"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/12\/11647259\/11599457.pdf?arnumber=11599457","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,12]],"date-time":"2026-08-12T19:03:51Z","timestamp":1786561431000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11599457\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,9]]},"references-count":52,"journal-issue":{"issue":"9"},"URL":"https:\/\/doi.org\/10.1109\/tc.2026.3710209","relation":{},"ISSN":["0018-9340","1557-9956","2326-3814"],"issn-type":[{"value":"0018-9340","type":"print"},{"value":"1557-9956","type":"electronic"},{"value":"2326-3814","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,9]]}}}