{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,2,20]],"date-time":"2026-02-20T07:06:40Z","timestamp":1771571200362,"version":"3.50.1"},"reference-count":37,"publisher":"IEEE","license":[{"start":{"date-parts":[[2025,10,12]],"date-time":"2025-10-12T00:00:00Z","timestamp":1760227200000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2025,10,12]],"date-time":"2025-10-12T00:00:00Z","timestamp":1760227200000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"DOI":"10.13039\/501100003725","name":"National Research Foundation of Korea","doi-asserted-by":"publisher","award":["2020M3H2A1078119"],"award-info":[{"award-number":["2020M3H2A1078119"]}],"id":[{"id":"10.13039\/501100003725","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100003052","name":"Ministry of Trade, Industry & Energy","doi-asserted-by":"publisher","id":[{"id":"10.13039\/501100003052","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2025,10,12]]},"DOI":"10.1109\/apccas67402.2025.11378253","type":"proceedings-article","created":{"date-parts":[[2026,2,19]],"date-time":"2026-02-19T20:54:40Z","timestamp":1771534480000},"page":"1-5","source":"Crossref","is-referenced-by-count":0,"title":["RoPE on the Fast Track: Latency-Optimized LLM Inference on GPUs with Low-Rank QKV Factorization and Rotary Positional Embedding"],"prefix":"10.1109","author":[{"given":"Gilhyeon","family":"Lee","sequence":"first","affiliation":[{"name":"Seoul National University of Science and Technology,Department of Electrical and Information Engineering and Research Center for Electrical and Information Technology,Seoul,Korea"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Jihoon","family":"Jang","sequence":"additional","affiliation":[{"name":"Seoul National University of Science and Technology,Department of Electrical and Information Engineering and Research Center for Electrical and Information Technology,Seoul,Korea"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Kyungmin","family":"Goh","sequence":"additional","affiliation":[{"name":"Seoul National University of Science and Technology,Department of Electrical and Information Engineering and Research Center for Electrical and Information Technology,Seoul,Korea"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Hyun","family":"Kim","sequence":"additional","affiliation":[{"name":"Seoul National University of Science and Technology,Department of Electrical and Information Engineering and Research Center for Electrical and Information Technology,Seoul,Korea"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"263","reference":[{"key":"ref1","article-title":"Gpt-4 technical report","author":"Achiam","year":"2023","journal-title":"arXiv preprint"},{"key":"ref2","article-title":"Llama 2: Open foundation and fine-tuned chat models","author":"Touvron","year":"2023","journal-title":"arXiv preprint"},{"key":"ref3","article-title":"The llama 3 herd of models","author":"Grattafiori","year":"2024","journal-title":"arXiv preprint"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.48550\/ARXIV.1706.03762"},{"key":"ref5","article-title":"Rethinking positional encoding in language pre-training","author":"Ke","year":"2020","journal-title":"arXiv preprint"},{"key":"ref6","article-title":"Train short, test long: Attention with linear biases enables input length extrapolation","author":"Press","year":"2021","journal-title":"arXiv preprint"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/P19-1285"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1016\/j.neucom.2023.127063"},{"key":"ref9","volume-title":"Mistral 7b","author":"Jiang","year":"2023"},{"key":"ref10","volume-title":"GPT-J-6B: A 6 Billion Parameter Autoregressive Language Model","author":"Wang","year":"2021"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2022.bigscience-1.9"},{"key":"ref12","article-title":"Qwen technical report","author":"Bai","year":"2023","journal-title":"arXiv preprint"},{"key":"ref13","article-title":"Deepseek-v3 technical report","author":"Liu","year":"2024","journal-title":"arXiv preprint"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.1109\/LCA.2025.3535470"},{"key":"ref15","article-title":"Powerinfer-2: Fast large language model inference on a smartphone","author":"Xue","year":"2024","journal-title":"arXiv preprint"},{"key":"ref16","first-page":"30318","article-title":"Gpt3. int8 (): 8-bit matrix multiplication for transformers at scale","volume":"35","author":"Dettmers","year":"2022","journal-title":"Advances in neural information processing systems"},{"key":"ref17","volume-title":"Llama 3.1: An in-depth analysis of the next-generation large language model","author":"Vavekanand","year":"2024"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1109\/TCSI.2025.3573274"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.1016\/j.sysarc.2024.103247"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.1007\/s11554-024-01442-8"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.1145\/3656177"},{"key":"ref22","article-title":"Longrope: Extending 11 m context window beyond 2 million tokens","author":"Ding","year":"2024","journal-title":"arXiv preprint"},{"key":"ref23","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2023.acl-long.816"},{"key":"ref24","article-title":"Yarn: Efficient context window extension of large language models","author":"Peng","year":"2023","journal-title":"arXiv preprint"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.1109\/ISCA52012.2021.00013"},{"key":"ref26","first-page":"11","volume-title":"NVIDIA Ada AD102 GPU Architecture Whitepaper","year":"2022"},{"key":"ref27","volume-title":"Nvidia nsight compute","year":"2024"},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.1016\/j.sysarc.2023.102888"},{"key":"ref29","doi-asserted-by":"publisher","DOI":"10.1109\/CGO.2009.20"},{"key":"ref30","volume-title":"CUDA C++ Programming Guide, section 6.2.8.2: Concurrent kernel execution","year":"2025"},{"key":"ref31","doi-asserted-by":"publisher","DOI":"10.1145\/3453953.3453972"},{"key":"ref32","volume-title":"NVIDIA Ampere GA102 GPU Architecture Whitepaper","year":"2020"},{"key":"ref33","doi-asserted-by":"publisher","DOI":"10.1145\/3315508.3329973"},{"key":"ref34","volume-title":"Intel core i9-10900x xseries processor (19.25m cache, 3.70 ghz) product specifications","year":"2019"},{"key":"ref35","doi-asserted-by":"publisher","DOI":"10.1145\/3676536.3676766"},{"key":"ref36","first-page":"80268037","article-title":"Pytorch: An imperative style, high-performance deep learning library","author":"Paszke","year":"2019","journal-title":"Advances in Neural Information Processing Systems (NeurIPS)"},{"key":"ref37","article-title":"Pointer sentinel mixture models","author":"Merity","year":"2016","journal-title":"arXiv preprint"}],"event":{"name":"2025 IEEE Asia Pacific Conference on Circuits and Systems (APCCAS)","location":"Busan, Korea, Republic of","start":{"date-parts":[[2025,10,12]]},"end":{"date-parts":[[2025,10,15]]}},"container-title":["2025 IEEE Asia Pacific Conference on Circuits and Systems (APCCAS)"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/11373921\/11376610\/11378253.pdf?arnumber=11378253","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,2,20]],"date-time":"2026-02-20T06:38:42Z","timestamp":1771569522000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11378253\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,10,12]]},"references-count":37,"URL":"https:\/\/doi.org\/10.1109\/apccas67402.2025.11378253","relation":{},"subject":[],"published":{"date-parts":[[2025,10,12]]}}}