Accelerating LLM Inference via Dynamic KV Cache Placement in Heterogeneous Memory System
IEEE Computer Architecture Letters · 2025-07 · 24 (2) · 337-340
DOI: 10.1109/lca.2025.3622724 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Computer Architecture Letters, 2025-07. Verfasst von Yunhua Fang und 8 weiteren Autor:innen. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 9 Literaturangaben.
Abstract (Originalsprache)
Large Language Model (LLM) inference is increasingly constrained by memory bandwidth, with frequent access to the key-value (KV) cache dominating data movement. While attention sparsity reduces some memory traffic, the relevance of past tokens varies over time, requiring the full KV cache to remain accessible and sustaining pressure on both bandwidth and capacity. With advances in interconnects such as NVLink and LPDDR5X, modern AI hardware now integrates high-bandwidth memory (HBM) with high-speed off-package DRAM, making heterogeneous memory systems a practical solution. This work investigates dynamic KV cache placement across such systems to maximize aggregated bandwidth utilization under capacity constraints. Rather than proposing a specific scheduling policy, we formulate the placement problem mathematically and derive a theoretical upper bound, revealing substantial headroom for runtime optimization. To our knowledge, this is the first formal treatment of dynamic KV cache scheduling in heterogeneous memory systems for LLM inference.
Zitieren
APA
Fang, Y., Xie, R., Haq, A. U., Ma, L., Maghraoui, K. E., Wang, N., Wang, M., Liu, L., & Zhang, T. (2025). Accelerating LLM Inference via Dynamic KV Cache Placement in Heterogeneous Memory System. IEEE Computer Architecture Letters, 24(2), 337-340. https://doi.org/10.1109/lca.2025.3622724
Vancouver
Fang Y, Xie R, Haq AU, Ma L, Maghraoui KE, Wang N, et al. Accelerating LLM Inference via Dynamic KV Cache Placement in Heterogeneous Memory System. IEEE Computer Architecture Letters. 2025;24(2):337-340. doi:10.1109/lca.2025.3622724
BibTeX
@article{Fang2025,
author = {Fang, Yunhua and Xie, Rui and Haq, Asad Ul and Ma, Linsen and Maghraoui, Kaoutar El and Wang, Naigang and Wang, Meng and Liu, Liu and Zhang, Tong},
title = {Accelerating LLM Inference via Dynamic KV Cache Placement in Heterogeneous Memory System},
journal = {IEEE Computer Architecture Letters},
year = {2025},
volume = {24},
number = {2},
pages = {337--340},
doi = {10.1109/lca.2025.3622724}
}
EUVIMED