PiKV: KV Cache Management System for Mixture of Experts

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Liu, Dong, Yu, Yanxuan, Lengerich, Ben, Wu, Ying Nian
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866911697747312640
author Liu, Dong
Yu, Yanxuan
Lengerich, Ben
Wu, Ying Nian
author_facet Liu, Dong
Yu, Yanxuan
Lengerich, Ben
Wu, Ying Nian
contents As large-scale language models continue to scale up in both size and context length, the memory and communication cost of key-value (KV) cache storage has become a major bottleneck in multi-GPU and multi-node inference. While MoE-based architectures sparsify computation across experts, the corresponding KV caches remain dense and globally synchronized, resulting in significant overhead. We introduce \textbf{PiKV}, a parallel and distributed KV cache serving framework tailored for MoE architecture. PiKV leverages \textit{expert-sharded KV storage} to partition caches across GPUs, \textit{PiKV routing} to reduce token-to-KV access, and a \textit{PiKV Scheduling} to adaptively retain query-relevant entries. To further reduce memory usage, PiKV integrates \textit{PiKV Compression} modules the caching pipeline for acceleration. PiKV is recently publicly available as an open-source software library: \href{https://github.com/NoakLiu/PiKV}{https://github.com/NoakLiu/PiKV}. PiKV is still a living project, aiming to become a comprehesive KV Cache management system for MoE Architectures.
format Preprint
id arxiv_https___arxiv_org_abs_2508_06526
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle PiKV: KV Cache Management System for Mixture of Experts
Liu, Dong
Yu, Yanxuan
Lengerich, Ben
Wu, Ying Nian
Distributed, Parallel, and Cluster Computing
Artificial Intelligence
Hardware Architecture
As large-scale language models continue to scale up in both size and context length, the memory and communication cost of key-value (KV) cache storage has become a major bottleneck in multi-GPU and multi-node inference. While MoE-based architectures sparsify computation across experts, the corresponding KV caches remain dense and globally synchronized, resulting in significant overhead. We introduce \textbf{PiKV}, a parallel and distributed KV cache serving framework tailored for MoE architecture. PiKV leverages \textit{expert-sharded KV storage} to partition caches across GPUs, \textit{PiKV routing} to reduce token-to-KV access, and a \textit{PiKV Scheduling} to adaptively retain query-relevant entries. To further reduce memory usage, PiKV integrates \textit{PiKV Compression} modules the caching pipeline for acceleration. PiKV is recently publicly available as an open-source software library: \href{https://github.com/NoakLiu/PiKV}{https://github.com/NoakLiu/PiKV}. PiKV is still a living project, aiming to become a comprehesive KV Cache management system for MoE Architectures.
title PiKV: KV Cache Management System for Mixture of Experts
topic Distributed, Parallel, and Cluster Computing
Artificial Intelligence
Hardware Architecture
url https://arxiv.org/abs/2508.06526