{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,1]],"date-time":"2026-05-01T16:04:18Z","timestamp":1777651458108,"version":"3.51.4"},"reference-count":50,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","issue":"4","license":[{"start":{"date-parts":[[2025,4,1]],"date-time":"2025-04-01T00:00:00Z","timestamp":1743465600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2025,4,1]],"date-time":"2025-04-01T00:00:00Z","timestamp":1743465600000},"content-version":"am","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2025,4,1]],"date-time":"2025-04-01T00:00:00Z","timestamp":1743465600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2025,4,1]],"date-time":"2025-04-01T00:00:00Z","timestamp":1743465600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"name":"SRC CoCoSys, AI Chip Center for Emerging Smart Systems (ACCESS), Hong Kong, SAR, SRC AIHW, Precourt, Samsung, Apple, NSF FuSe, DARPA 3DSoC, SystemX Alliance, SGF, and Silitronics"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE J. Solid-State Circuits"],"published-print":{"date-parts":[[2025,4]]},"DOI":"10.1109\/jssc.2025.3545731","type":"journal-article","created":{"date-parts":[[2025,3,6]],"date-time":"2025-03-06T13:41:33Z","timestamp":1741268493000},"page":"1311-1323","source":"Crossref","is-referenced-by-count":10,"title":["MINOTAUR: A Posit-Based 0.42\u20130.50-TOPS\/W Edge Transformer Inference and Training Accelerator"],"prefix":"10.1109","volume":"60","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-4179-1692","authenticated-orcid":false,"given":"Kartik","family":"Prabhu","sequence":"first","affiliation":[{"name":"Department of Electrical Engineering, Stanford University, Stanford, CA, USA"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-3393-5489","authenticated-orcid":false,"given":"Robert M.","family":"Radway","sequence":"additional","affiliation":[{"name":"Department of Electrical Engineering, Stanford University, Stanford, CA, USA"}]},{"given":"Jeffrey","family":"Yu","sequence":"additional","affiliation":[{"name":"Department of Electrical Engineering, Stanford University, Stanford, CA, USA"}]},{"given":"Kai","family":"Bartolone","sequence":"additional","affiliation":[{"name":"Department of Electrical Engineering, Stanford University, Stanford, CA, USA"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-7012-4135","authenticated-orcid":false,"given":"Massimo","family":"Giordano","sequence":"additional","affiliation":[{"name":"Department of Electrical Engineering, Stanford University, Stanford, CA, USA"}]},{"given":"Fabian","family":"Peddinghaus","sequence":"additional","affiliation":[{"name":"Department of Electrical Engineering, Stanford University, Stanford, CA, USA"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-5763-8174","authenticated-orcid":false,"given":"Yonatan","family":"Urman","sequence":"additional","affiliation":[{"name":"Department of Electrical Engineering, Stanford University, Stanford, CA, USA"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-6283-3564","authenticated-orcid":false,"given":"Win-San","family":"Khwa","sequence":"additional","affiliation":[{"name":"Taiwan Semiconductor Manufacturing Company (TSMC), Hsinchu, Taiwan"}]},{"given":"Yu-Der","family":"Chih","sequence":"additional","affiliation":[{"name":"Taiwan Semiconductor Manufacturing Company (TSMC), Hsinchu, Taiwan"}]},{"given":"Meng-Fan","family":"Chang","sequence":"additional","affiliation":[{"name":"Taiwan Semiconductor Manufacturing Company (TSMC), Hsinchu, Taiwan"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-5572-5194","authenticated-orcid":false,"given":"Subhasish","family":"Mitra","sequence":"additional","affiliation":[{"name":"Department of Electrical Engineering, Stanford University, Stanford, CA, USA"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-8834-8663","authenticated-orcid":false,"given":"Priyanka","family":"Raina","sequence":"additional","affiliation":[{"name":"Department of Electrical Engineering, Stanford University, Stanford, CA, USA"}]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.48550\/ARXIV.1706.03762"},{"key":"ref2","article-title":"BERT: Pre-training of deep bidirectional transformers for language understanding","author":"Devlin","year":"2018","journal-title":"arXiv:1810.04805"},{"issue":"1","key":"ref3","first-page":"5485","article-title":"Exploring the limits of transfer learning with a unified text-to-text transformer","volume":"21","author":"Raffel","year":"2020","journal-title":"J. Mach. Learn. Res."},{"key":"ref4","article-title":"GPT-4 technical report","volume-title":"arXiv:2303.08774","author":"OpenAI","year":"2023"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.14529\/jsfi170206"},{"key":"ref6","first-page":"1","article-title":"LoRA: Low-rank adaptation of large language models","volume-title":"Proc. Int. Conf. Learn. Represent.","author":"Hu"},{"key":"ref7","article-title":"FP8 formats for deep learning","author":"Micikevicius","year":"2022","journal-title":"arXiv:2209.05433"},{"key":"ref8","first-page":"30318","article-title":"LLM.int8(): 8-bit matrix multiplication for transformers at scale","volume-title":"Proc. 36th Int. Conf. Neural Inf. Process. Syst","author":"Dettmers"},{"key":"ref9","article-title":"GPTQ: Accurate post-training quantization for generative pre-trained transformers","author":"Frantar","year":"2022","journal-title":"arXiv:2210.17323"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.1145\/3620666.3651368"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.3390\/s20051515"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.1109\/dac18074.2021.9586134"},{"key":"ref13","article-title":"The rocket chip generator","author":"Asanovic","year":"2016"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.1109\/MM.2020.2996616"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.1145\/3489517.3530633"},{"key":"ref16","first-page":"342","article-title":"A 12 nm 18.1TFLOPs\/W sparse transformer processor with entropy-based early exit, mixed-precision predication and fine-grained power management","volume-title":"IEEE Int. Solid-State Circuits Conf. (ISSCC) Dig. Tech. Papers","author":"Tambe"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1109\/VLSITechnologyandCir46769.2022.9830277"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1109\/hpca47549.2020.00035"},{"key":"ref19","first-page":"478","article-title":"An N40 256Kx44 embedded RRAM macro with SL-precharge SA and low-voltage current limiter to improve read and write performance","volume-title":"IEEE Int. Solid-State Circuits Conf. (ISSCC) Dig. Tech. Papers","author":"Chou"},{"key":"ref20","volume-title":"Wikimedia Foundation","year":"2024"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.1109\/JSSC.2022.3140753"},{"key":"ref22","article-title":"Training deep nets with sublinear memory cost","author":"Chen","year":"2016","journal-title":"arXiv:1604.06174"},{"key":"ref23","first-page":"1742","article-title":"Flexpoint: An adaptive numerical format for efficient training of deep neural networks","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"30","author":"K\u00f6ster"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.1109\/DAC18072.2020.9218516"},{"key":"ref25","article-title":"QLoRA: Efficient finetuning of quantized LLMs","author":"Dettmers","year":"2023","journal-title":"arXiv:2305.14314"},{"key":"ref26","article-title":"Extreme compression of large language models via additive quantization","author":"Egiazarian","year":"2024","journal-title":"arXiv:2401.06118"},{"key":"ref27","article-title":"BitNet: Scaling 1-bit transformers for large language models","author":"Wang","year":"2023","journal-title":"arXiv:2310.11453"},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.1109\/emc2-nips53020.2019.00016"},{"key":"ref29","article-title":"Microscaling data formats for deep learning","author":"Rouhani","year":"2023","journal-title":"arXiv:2310.10537"},{"key":"ref30","article-title":"Generating long sequences with sparse transformers","author":"Child","year":"2019","journal-title":"arXiv:1904.10509"},{"key":"ref31","article-title":"MiniLM: Deep self-attention distillation for task-agnostic compression of pre-trained transformers","author":"Wang","year":"2020","journal-title":"arXiv:2002.10957"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.1109\/isocc53507.2021.9613933"},{"key":"ref33","article-title":"FlashAttention: Fast and memory-efficient exact attention with IO-awareness","author":"Dao","year":"2022","journal-title":"arXiv:2205.14135"},{"key":"ref34","doi-asserted-by":"publisher","DOI":"10.1109\/vlsitechnologyandcir46769.2022.9830226"},{"key":"ref35","doi-asserted-by":"publisher","DOI":"10.1109\/VLSITechnologyandCir46769.2022.9830409"},{"key":"ref36","first-page":"72","article-title":"A 22 nm 3.5TOPS\/W flexible micro-robotic vision SoC with 2MB eMRAM for fully-on-chip intelligence","volume-title":"Proc. IEEE Symp. VLSI Technol. Circuits","author":"Zhang"},{"key":"ref37","first-page":"25","article-title":"A 22 nm 9.51 TOPS\/W neural engine with 2MB MRAM leveraging sparse-orthogonal Walsh-Hadamard transform computations and dynamic power gating","volume-title":"Proc. IEEE Eur. Solid-State Electron. Res. Conf. (ESSERC)","author":"Abillama"},{"key":"ref38","doi-asserted-by":"publisher","DOI":"10.1109\/JSSC.2024.3387566"},{"key":"ref39","doi-asserted-by":"publisher","DOI":"10.1109\/VLSITechnologyandCir46769.2022.9830487"},{"key":"ref40","first-page":"1","article-title":"A 28 nm 16.9\u2013300TOPS\/W computing-in-memory processor supporting floating-point NN inference\/training with intensive-CIM sparse-digital architecture","volume-title":"Proc. IEEE Int. Solid-State Circuits Conf. (ISSCC)","author":"Yue"},{"key":"ref41","article-title":"CAMEL: Co-designing AI models and embedded DRAMs for efficient on-device learning","author":"Qian Zhang","year":"2023","journal-title":"arXiv:2305.03148"},{"key":"ref42","volume-title":"NVIDIA Jetson Orin","year":"2025"},{"key":"ref43","doi-asserted-by":"publisher","DOI":"10.1109\/hotchips.2019.8875657"},{"key":"ref44","doi-asserted-by":"publisher","DOI":"10.1109\/JSSC.2023.3234893"},{"key":"ref45","doi-asserted-by":"publisher","DOI":"10.23919\/VLSICircuits52068.2021.9492347"},{"key":"ref46","doi-asserted-by":"publisher","DOI":"10.23919\/VLSIC.2019.8778056"},{"key":"ref47","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/w18-5446"},{"key":"ref48","article-title":"MobileBERT: A compact task-agnostic BERT for resource-limited devices","author":"Sun","year":"2020","journal-title":"arXiv:2004.02984"},{"key":"ref49","volume-title":"TorchVision: PyTorch\u2019s Computer Vision Library"},{"key":"ref50","first-page":"9","article-title":"Automatically constructing a corpus of sentential paraphrases","volume-title":"Proc. 3rd Int. Workshop Paraphrasing (IWP)","author":"Dolan"}],"container-title":["IEEE Journal of Solid-State Circuits"],"original-title":[],"link":[{"URL":"https:\/\/ieeexplore.ieee.org\/ielam\/4\/10944491\/10916649-aam.pdf","content-type":"application\/pdf","content-version":"am","intended-application":"syndication"},{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/4\/10944491\/10916649.pdf?arnumber=10916649","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,12,2]],"date-time":"2025-12-02T18:48:21Z","timestamp":1764701301000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/10916649\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,4]]},"references-count":50,"journal-issue":{"issue":"4"},"URL":"https:\/\/doi.org\/10.1109\/jssc.2025.3545731","relation":{},"ISSN":["0018-9200","1558-173X"],"issn-type":[{"value":"0018-9200","type":"print"},{"value":"1558-173X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,4]]}}}