{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,3,10]],"date-time":"2026-03-10T14:53:18Z","timestamp":1773154398111,"version":"3.50.1"},"reference-count":38,"publisher":"IEEE","license":[{"start":{"date-parts":[[2025,3,31]],"date-time":"2025-03-31T00:00:00Z","timestamp":1743379200000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2025,3,31]],"date-time":"2025-03-31T00:00:00Z","timestamp":1743379200000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"DOI":"10.13039\/501100011178","name":"Institute of Automation, Chinese Academy of Sciences","doi-asserted-by":"publisher","award":["E411230101"],"award-info":[{"award-number":["E411230101"]}],"id":[{"id":"10.13039\/501100011178","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2025,3,31]]},"DOI":"10.23919\/date64628.2025.10993087","type":"proceedings-article","created":{"date-parts":[[2025,5,21]],"date-time":"2025-05-21T17:36:35Z","timestamp":1747848995000},"page":"1-7","source":"Crossref","is-referenced-by-count":7,"title":["Pushing up to the Limit of Memory Bandwidth and Capacity Utilization for Efficient LLM Decoding on Embedded FPGA"],"prefix":"10.23919","author":[{"given":"Jindong","family":"Li","sequence":"first","affiliation":[{"name":"Institute of Automation, Chinese Academy of Sciences,Brain-inspired Cognitive Intelligence Lab"}]},{"given":"Tenglong","family":"Li","sequence":"additional","affiliation":[{"name":"Institute of Automation, Chinese Academy of Sciences,Brain-inspired Cognitive Intelligence Lab"}]},{"given":"Guobin","family":"Shen","sequence":"additional","affiliation":[{"name":"Institute of Automation, Chinese Academy of Sciences,Brain-inspired Cognitive Intelligence Lab"}]},{"given":"Dongcheng","family":"Zhao","sequence":"additional","affiliation":[{"name":"Institute of Automation, Chinese Academy of Sciences,Brain-inspired Cognitive Intelligence Lab"}]},{"given":"Qian","family":"Zhang","sequence":"additional","affiliation":[{"name":"Institute of Automation, Chinese Academy of Sciences,Brain-inspired Cognitive Intelligence Lab"}]},{"given":"Yi","family":"Zeng","sequence":"additional","affiliation":[{"name":"Institute of Automation, Chinese Academy of Sciences,Brain-inspired Cognitive Intelligence Lab"}]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.1109\/TCAD.2017.2705069"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.1109\/ICTA53157.2021.9661886"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.1109\/TCAD.2024.3380550"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.1109\/TCSI.2024.3496554"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1109\/FPL57034.2022.00027"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.1109\/HPCA56546.2023.10071047"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1109\/MICRO56248.2022.00051"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1145\/3626202.3637562"},{"key":"ref9","article-title":"Llama 2: Open foundation and finetuned chat models","author":"Touvron","year":"2023","journal-title":"arXiv preprint"},{"key":"ref10","article-title":"Root mean square layer normalization","volume":"32","author":"Zhang","year":"2019","journal-title":"Advances in Neural Information Processing Systems"},{"key":"ref11","article-title":"Online normalizer calculation for softmax","author":"Milakov","year":"2018","journal-title":"arXiv preprint"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.1016\/j.neucom.2023.127063"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1016\/j.neunet.2017.12.012"},{"key":"ref14","article-title":"Attention is all you need","author":"Vaswani","year":"2017","journal-title":"Advances in Neural Information Processing Systems"},{"key":"ref15","article-title":"An image is worth 16\u00d716 words: Transformers for image recognition at scale","author":"Dosovitskiy","year":"2020","journal-title":"arXiv preprint"},{"key":"ref16","article-title":"Bert: Pre-training of deep bidirectional transformers for language understanding","author":"Devlin","year":"2018","journal-title":"arXiv preprint"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.23919\/DATE51398.2021.9474043"},{"issue":"8","key":"ref18","first-page":"9","article-title":"Language models are unsupervised multitask learners","volume":"1","author":"Radford","year":"2019","journal-title":"OpenAI blog"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.1145\/3656177"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.1109\/tcsi.2025.3546256"},{"key":"ref21","first-page":"38 087","article-title":"Smoothquant: Accurate and efficient post-training quantization for large language models","volume-title":"International Conference on Machine Learning","author":"Xiao","year":"2023"},{"key":"ref22","first-page":"87","article-title":"Awq: Activation-aware weight quantization for on-device 11m compression and acceleration","volume":"6","author":"Lin","year":"2024","journal-title":"Proceedings of Machine Learning and Systems"},{"key":"ref23","article-title":"Evaluating quantized large language models","author":"Li","year":"2024","journal-title":"arXiv preprint"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.1109\/FPL64840.2024.00035"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.1109\/TVLSI.2023.3279349"},{"key":"ref26","article-title":"I-llm: Efficient integer-only inference for fully-quantized low-bit large language models","author":"Hu","year":"2024","journal-title":"arXiv preprint"},{"key":"ref27","article-title":"Spinalhdl: Scala based hdl"},{"key":"ref28","article-title":"Cocotb: a python-based digital logic verification framework","author":"Rosser","year":"2018","journal-title":"Micro-electronics Section seminar. CERN, Geneva, Switzerland"},{"key":"ref29","article-title":"Designing efficient llm accelerators for edge devices","author":"Haris","year":"2024","journal-title":"arXiv preprint"},{"key":"ref30","doi-asserted-by":"crossref","DOI":"10.1109\/WF-IoT62078.2024.10811385","article-title":"Llamaf: An efficient llama2 architecture accelerator on embedded fpgas","author":"Xu","year":"2024"},{"key":"ref31","article-title":"Georgi gerganov. ggerganov\/llama.cpp: Port of facebook\u2019s llama model in c\/c++"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.1145\/3603287.3651205"},{"key":"ref33","article-title":"Tinychat: Large language model on the edge"},{"key":"ref34","article-title":"Tinychat: Efficient and lightweight chatbot with awq"},{"key":"ref35","article-title":"Optimized local inference for llms with huggingface-like apis for quantization, vision\/language models, multimodal agents, speech, vector db, and rag"},{"key":"ref36","article-title":"Jetson ai lab benchmark"},{"key":"ref37","article-title":"Braincog: Brain-inspired cognitive intelligence engine"},{"key":"ref38","doi-asserted-by":"publisher","DOI":"10.1016\/j.patter.2023.100789"}],"event":{"name":"2025 Design, Automation & Test in Europe Conference (DATE)","location":"Lyon, France","start":{"date-parts":[[2025,3,31]]},"end":{"date-parts":[[2025,4,2]]}},"container-title":["2025 Design, Automation & Test in Europe Conference (DATE)"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/10992638\/10992588\/10993087.pdf?arnumber=10993087","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,5,22]],"date-time":"2025-05-22T05:33:06Z","timestamp":1747891986000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/10993087\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,3,31]]},"references-count":38,"URL":"https:\/\/doi.org\/10.23919\/date64628.2025.10993087","relation":{},"subject":[],"published":{"date-parts":[[2025,3,31]]}}}