# Towards Efficient Key-Value Cache Management for Prefix Prefilling in LLM Inference

arXiv: https://arxiv.org/abs/2505.21919 · HTML: https://mlsanity.com/paper/2505.21919/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Placed in KV cache optimization, one of 12 directions in Efficient LLM inference and serving.
- Builds on 6 of its 14 references that are in the map.
- First posted on arXiv 28 May 2025.

## Placed in

- Efficient LLM inference and serving / [KV cache optimization](https://mlsanity.com/inference/kv-cache-optimization/)

## Builds on (most cited first, 6 of 6)

- [Longformer: The Long-Document Transformer](https://mlsanity.com/paper/2004.05150/) (2020)
- [GQA: Training Generalized Multi-Query Transformer Models from Multi-Head Checkpoints](https://mlsanity.com/paper/2305.13245/) (2023)
- [CacheGen: KV Cache Compression and Streaming for Fast Large Language Model Serving](https://mlsanity.com/paper/2310.07240/) (2023)
- [Pie: Pooling CPU Memory for LLM Inference](https://mlsanity.com/paper/2411.09317/) (2024)
- [More Tokens, Lower Precision: Towards the Optimal Token-Precision Trade-off in KV Cache Compression](https://mlsanity.com/paper/2412.12706/) (2024)
- [Sherman: A Write-Optimized Distributed B+Tree Index on Disaggregated Memory](https://mlsanity.com/paper/2112.07320/) (2021)

## Built on by (most cited first, 2 of 2)

- [A Survey of Context Engineering for Large Language Models](https://mlsanity.com/paper/2507.13334/) (2025)
- [DeepSieve: Information Sieving via LLM-as-a-Knowledge-Router](https://mlsanity.com/paper/2507.22050/) (2025)

## Abstract (arXiv)

The increasing adoption of large language models (LLMs) with extended context windows necessitates efficient Key-Value Cache (KVC) management to optimize inference performance. Inference workloads like Retrieval-Augmented Generation (RAG) and agents exhibit high cache reusability, making efficient caching critical to reducing redundancy and improving speed. We analyze real-world KVC access patterns using publicly available traces and evaluate commercial key-value stores like Redis and state-of-the-art RDMA-based systems (CHIME [1] and Sherman [2]) for KVC metadata management. Our work demonstrates the lack of tailored storage solution for KVC prefilling, underscores the need for an efficient distributed caching system with optimized metadata management for LLM workloads, and provides insights into designing improved KVC management systems for scalable, low-latency inference.