{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,1]],"date-time":"2026-07-01T20:40:57Z","timestamp":1782938457898,"version":"3.54.5"},"reference-count":43,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","issue":"7","license":[{"start":{"date-parts":[[2026,7,1]],"date-time":"2026-07-01T00:00:00Z","timestamp":1782864000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2026,7,1]],"date-time":"2026-07-01T00:00:00Z","timestamp":1782864000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2026,7,1]],"date-time":"2026-07-01T00:00:00Z","timestamp":1782864000000},"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":["IEEE Trans. VLSI Syst."],"published-print":{"date-parts":[[2026,7]]},"DOI":"10.1109\/tvlsi.2026.3686720","type":"journal-article","created":{"date-parts":[[2026,5,5]],"date-time":"2026-05-05T20:05:44Z","timestamp":1778011544000},"page":"2176-2186","source":"Crossref","is-referenced-by-count":0,"title":["Algorithm--Hardware Co-Design of Digital Compute-in-Memory Architecture Supporting Flexible and Temporal N:M Sparsity"],"prefix":"10.1109","volume":"34","author":[{"ORCID":"https:\/\/orcid.org\/0009-0000-4763-3321","authenticated-orcid":false,"given":"Akshat","family":"Ramachandran","sequence":"first","affiliation":[{"name":"Georgia Institute of Technology, Atlanta, GA, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Souvik","family":"Kundu","sequence":"additional","affiliation":[{"name":"Intel Corporation, Santa Clara, CA, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-8848-1069","authenticated-orcid":false,"given":"Arnab","family":"Raha","sequence":"additional","affiliation":[{"name":"Intel Corporation, Santa Clara, CA, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-5992-8554","authenticated-orcid":false,"given":"Shamik","family":"Kundu","sequence":"additional","affiliation":[{"name":"Intel Corporation, Santa Clara, CA, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0004-3768-1337","authenticated-orcid":false,"given":"Deepak A.","family":"Mathaikutty","sequence":"additional","affiliation":[{"name":"Intel Corporation, Santa Clara, CA, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-5738-6942","authenticated-orcid":false,"given":"Tushar","family":"Krishna","sequence":"additional","affiliation":[{"name":"Georgia Institute of Technology, Atlanta, GA, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.1109\/ISSCC49663.2026.11409207"},{"key":"ref2","article-title":"OpenFlamingo: An open-source framework for training large autoregressive vision-language models","author":"Awadalla","year":"2023","journal-title":"arXiv:2308.01390"},{"key":"ref3","first-page":"21437","article-title":"Thinking clearly: Enhancing reasoning with redundant token pruning","volume-title":"Proc. Findings EMNLP","author":"Choi"},{"key":"ref4","article-title":"HGCA: Hybrid GPU-CPU attention for long context LLM inference","author":"Deng","year":"2025","journal-title":"arXiv:2507.03153"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1109\/EMC249363.2019.00012"},{"key":"ref6","first-page":"10323","article-title":"SparseGPT: Massive language models can be accurately pruned in one-shot","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Frantar"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1109\/ISSCC42614.2022.9731754"},{"key":"ref8","article-title":"The pile: An 800GB dataset of diverse text for language modeling","author":"Gao","year":"2021","journal-title":"arXiv:2101.00027"},{"key":"ref9","article-title":"Mamba: Linear-time sequence modeling with selective state spaces","author":"Gu","year":"2023","journal-title":"arXiv:2312.00752"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.1109\/HPCA56546.2023.10071058"},{"key":"ref11","article-title":"Mixtral of experts","author":"Jiang","year":"2024","journal-title":"arXiv:2401.04088"},{"key":"ref12","article-title":"Recap: Training-free compensation for coarse activation channel pruning in compressed LLMs","author":"Lee","year":"2025","journal-title":"Mach. Learn. Comput. Archit. Syst."},{"key":"ref13","article-title":"Polestar-cache: Reconciling parallel decoding and accuracy in diffusion LLMs via token drift-aware KV cache recalibration","volume-title":"Proc. Workshop Latent Implicit Thinking\u2013Going Beyond CoT Reasoning","author":"Lee"},{"key":"ref14","article-title":"Making slow thinking faster: Compressing LLM chain-of-thought via step entropy","author":"Li","year":"2025","journal-title":"arXiv:2508.03346"},{"key":"ref15","first-page":"500","article-title":"A fully integrated analog ReRAM based 78.4 TOPS\/W compute-in-memory chip with fully parallel MAC computing","volume-title":"IEEE Int. Solid-State Circuits Conf. (ISSCC) Dig. Tech. Papers","author":"Liu"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.1109\/HPCA53966.2022.00049"},{"key":"ref17","article-title":"Reasoning models can be accurately pruned via chain-of-thought reconstruction","author":"Lucas","year":"2025","journal-title":"arXiv:2509.12464"},{"key":"ref18","volume-title":"Aime 2024 Problems","year":"2024"},{"key":"ref19","volume-title":"Introducing Meta LLAMA 3: The Most Capable Openly Available Llm To Date","year":"2024"},{"key":"ref20","article-title":"Accelerating sparse deep neural networks","author":"Mishra","year":"2021","journal-title":"arXiv:2104.08378"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2025.emnlp-main.1025"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.1145\/3140659.3080254"},{"key":"ref23","doi-asserted-by":"publisher","DOI":"10.3389\/fhpcp.2025.1570210"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.1145\/3695053.3730989"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.1109\/ISLPED65674.2025.11261770"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51701.2025.01967"},{"key":"ref27","article-title":"ThinKV: Thought-adaptive KV cache compression for efficient reasoning models","author":"Ramachandran","year":"2025","journal-title":"arXiv:2510.01290"},{"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","doi-asserted-by":"publisher","DOI":"10.23919\/DATE64628.2025.10993035"},{"key":"ref31","article-title":"Dnnweaver: From high-level deep network models to FPGA acceleration","volume-title":"Proc. Workshop Cognit. Architectures","author":"Sharma"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.emnlp-main.68"},{"key":"ref33","doi-asserted-by":"publisher","DOI":"10.1109\/CICC60959.2024.10529009"},{"key":"ref34","doi-asserted-by":"publisher","DOI":"10.1109\/MM.2023.3338059"},{"key":"ref35","first-page":"4942","article-title":"A simple and effective pruning approach for large language models","volume-title":"Proc. ICLR","author":"Sun"},{"key":"ref36","article-title":"Llama 2: Open foundation and fine-tuned chat models","author":"Touvron","year":"2023","journal-title":"arXiv:2307.09288"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.1109\/TCAD.2022.3172600"},{"key":"ref38","first-page":"1106","article-title":"HighLight: Efficient and flexible DNN acceleration with hierarchical structured sparsity","volume-title":"Proc. 56th Annu. IEEE\/ACM Int. Symp. Microarchitecture","author":"Wu"},{"key":"ref39","article-title":"A task-centric angle of LLM pre-trained weights through sparsity","volume-title":"Proc. ICML","author":"Yin"},{"key":"ref40","article-title":"Outlier weighed layerwise sparsity (OWL): A missing secret sauce for pruning LLMS to high sparsity","volume-title":"Proc. ICML","author":"Yin"},{"key":"ref41","first-page":"238","article-title":"A 2.75-to-75.9 TOPS\/W computing-in-memory NN processor supporting set-associate block-wise zero skipping and ping-pong CIM with simultaneous computation and weight updating","volume-title":"IEEE Int. Solid-State Circuits Conf. (ISSCC) Dig. Tech. Papers","author":"Yue"},{"key":"ref42","doi-asserted-by":"publisher","DOI":"10.1109\/JSSC.2024.3387995"},{"key":"ref43","article-title":"Dynamic sparse no training: Training-free fine-tuning for sparse LLMs","author":"Zhang","year":"2023","journal-title":"arXiv:2310.08915"}],"container-title":["IEEE Transactions on Very Large Scale Integration (VLSI) Systems"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/92\/11589510\/11506321.pdf?arnumber=11506321","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,7,1]],"date-time":"2026-07-01T19:39:12Z","timestamp":1782934752000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11506321\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,7]]},"references-count":43,"journal-issue":{"issue":"7"},"URL":"https:\/\/doi.org\/10.1109\/tvlsi.2026.3686720","relation":{},"ISSN":["1063-8210","1557-9999"],"issn-type":[{"value":"1063-8210","type":"print"},{"value":"1557-9999","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,7]]}}}