{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,22]],"date-time":"2026-10-22T17:06:54Z","timestamp":1792688814635,"version":"4.5.0"},"reference-count":45,"publisher":"IEEE","license":[{"start":{"date-parts":[[2025,8,6]],"date-time":"2025-08-06T00:00:00Z","timestamp":1754438400000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2025,8,6]],"date-time":"2025-08-06T00:00:00Z","timestamp":1754438400000},"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":[[2025,8,6]]},"DOI":"10.1109\/islped65674.2025.11261770","type":"proceedings-article","created":{"date-parts":[[2025,12,3]],"date-time":"2025-12-03T18:39:13Z","timestamp":1764787153000},"page":"1-7","source":"Crossref","is-referenced-by-count":3,"title":["Accelerating LLM Inference with Flexible N:M Sparsity via A Fully Digital Compute-in-Memory Accelerator"],"prefix":"10.1109","author":[{"given":"Akshat","family":"Ramachandran","sequence":"first","affiliation":[{"name":"Georgia Institute of Technology,USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Souvik","family":"Kundu","sequence":"additional","affiliation":[{"name":"Intel Labs,USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Arnab","family":"Raha","sequence":"additional","affiliation":[{"name":"Intel Corporation,USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Shamik","family":"Kundu","sequence":"additional","affiliation":[{"name":"Intel Corporation,USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Deepak K.","family":"Mathaikutty","sequence":"additional","affiliation":[{"name":"Intel Corporation,USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Tushar","family":"Krishna","sequence":"additional","affiliation":[{"name":"Georgia Institute of Technology,USA"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","article-title":"Metron: Holistic performance evaluation framework for llm inference systems","author":"Agrawal","year":"2024"},{"key":"ref2","article-title":"Openflamingo: An open-source framework for training large autoregressive vision-language models","author":"Awadalla","year":"2023"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v34i05.6239"},{"key":"ref4","article-title":"Microsoft coco captions: Data collection and evaluation server","author":"Chen","year":"2015"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1109\/ISSCC42613.2021.9366045"},{"key":"ref6","article-title":"Boolq: Exploring the surprising difficulty of natural yes\/no questions","author":"Clark","year":"2019"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1145\/3649329.3655948"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1109\/EMC249363.2019.00012"},{"key":"ref9","article-title":"SparseGPT: Massive language models can be accurately pruned in one-shot","author":"Frantar","year":"2023"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.1109\/ISSCC42614.2022.9731754"},{"key":"ref11","article-title":"The pile: An 800gb dataset of diverse text for language modeling","author":"Gao","year":"2020"},{"key":"ref12","article-title":"Mamba: Linear-time sequence modeling with selective state spaces","author":"Gu","year":"2023"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1109\/ISCA52012.2021.00011"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.1109\/HPCA56546.2023.10071058"},{"key":"ref15","article-title":"Mixtral of experts","author":"Jiang","year":"2024"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.1109\/JSSC.2021.3061508"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1109\/ISLPED58423.2023.10244499"},{"key":"ref18","first-page":"87","article-title":"Awq: Activation-aware weight quantization for on-device llm compression and acceleration","volume-title":"Proceedings of Machine Learning and Systems","volume":"6","author":"Lin"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.1109\/ISSCC19947.2020.9062953"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.1109\/HPCA53966.2022.00049"},{"key":"ref21","article-title":"An analysis of neural language modeling at multiple scales","author":"Merity","year":"2018"},{"key":"ref22","article-title":"Introducing meta llama 3: The most capable openly available llm to date","author":"Meta","year":"2024","journal-title":"Meta AI"},{"key":"ref23","article-title":"Accelerating sparse deep neural networks","author":"Mishra","year":"2021"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.1145\/3140659.3080254"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.3389\/fhpcp.2025.1570210"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1145\/3695053.3730989"},{"key":"ref27","article-title":"Ouromamba: A data-free quantization framework for vision mamba models","author":"Ramachandran","year":"2025"},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.1145\/3649329.3656544"},{"key":"ref29","doi-asserted-by":"publisher","DOI":"10.1109\/ISCAS51556.2021.9401196"},{"key":"ref30","article-title":"Dnnweaver: From high-level deep network models to fpga acceleration","volume-title":"the Workshop on Cognitive Architectures","author":"Sharma"},{"key":"ref31","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00851"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.1109\/CICC60959.2024.10529009"},{"key":"ref33","doi-asserted-by":"publisher","DOI":"10.1109\/mm.2023.3338059"},{"key":"ref34","article-title":"A simple and effective pruning approach for large language models","volume-title":"ICLR","author":"Sun"},{"key":"ref35","article-title":"Llama 2: Open foundation and fine-tuned chat models","author":"Touvron","year":"2023"},{"key":"ref36","doi-asserted-by":"publisher","DOI":"10.1109\/TCAD.2022.3172600"},{"key":"ref37","first-page":"1106","article-title":"Highlight: Efficient and flexible dnn acceleration with hierarchical structured sparsity","volume-title":"Proceedings of the 56th Annual IEEE\/ACM International Symposium on Microarchitecture","author":"Wu"},{"key":"ref38","article-title":"A task-centric angle of llm pre-trained weights through sparsity","volume-title":"ICML","author":"Yin"},{"key":"ref39","article-title":"Outlier weighed layerwise sparsity: A missing secret sauce for pruning llms to high sparsity","volume-title":"ICML","author":"Yin"},{"key":"ref40","doi-asserted-by":"publisher","DOI":"10.1109\/MCAS.2021.3092533"},{"key":"ref41","doi-asserted-by":"publisher","DOI":"10.1109\/ISSCC42613.2021.9365958"},{"key":"ref42","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/P19-1472"},{"key":"ref43","doi-asserted-by":"publisher","DOI":"10.1109\/JSSC.2024.3387995"},{"key":"ref44","article-title":"Dynamic sparse no training: Training-free fine-tuning for sparse llms","author":"Zhang","year":"2023"},{"key":"ref45","doi-asserted-by":"publisher","DOI":"10.1109\/ISVLSI59464.2023.10238597"}],"event":{"name":"2025 IEEE\/ACM International Symposium on Low Power Electronics and Design (ISLPED)","location":"Reykjav\u00edk, Iceland","start":{"date-parts":[[2025,8,6]]},"end":{"date-parts":[[2025,8,8]]}},"container-title":["2025 IEEE\/ACM International Symposium on Low Power Electronics and Design (ISLPED)"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/11261694\/11261695\/11261770.pdf?arnumber=11261770","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,1,27]],"date-time":"2026-01-27T04:47:18Z","timestamp":1769489238000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11261770\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,8,6]]},"references-count":45,"URL":"https:\/\/doi.org\/10.1109\/islped65674.2025.11261770","relation":{},"subject":[],"published":{"date-parts":[[2025,8,6]]}}}