# Multi-Segment Attention: Enabling Efficient KV-Cache Management for Faster Large Language Model Serving

arXiv: https://arxiv.org/abs/2606.02964 · HTML: https://mlsanity.com/paper/2606.02964/

## Key facts

- Cited by 3 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 30 of its 56 references that are in the map.
- First posted on arXiv 1 Jun 2026.

## Placed in

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

## Builds on (most cited first, 30 of 30)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Efficient Memory Management for Large Language Model Serving with PagedAttention](https://mlsanity.com/paper/2309.06180/) (2023)
- [LongBench: A Bilingual, Multitask Benchmark for Long Context Understanding](https://mlsanity.com/paper/2308.14508/) (2023)
- [GQA: Training Generalized Multi-Query Transformer Models from Multi-Head Checkpoints](https://mlsanity.com/paper/2305.13245/) (2023)
- [SGLang: Efficient Execution of Structured Language Model Programs](https://mlsanity.com/paper/2312.07104/) (2023)
- [DeepSeek-V2: A Strong, Economical, and Efficient Mixture-of-Experts Language Model](https://mlsanity.com/paper/2405.04434/) (2024)
- [Taming Throughput-Latency Tradeoff in LLM Inference with Sarathi-Serve](https://mlsanity.com/paper/2403.02310/) (2024)
- [CacheGen: KV Cache Compression and Streaming for Fast Large Language Model Serving](https://mlsanity.com/paper/2310.07240/) (2023)
- [LooGLE: Can Long-Context Language Models Understand Long Contexts?](https://mlsanity.com/paper/2311.04939/) (2023)
- [Cost-Efficient Large Language Model Serving for Multi-turn Conversations with CachedAttention](https://mlsanity.com/paper/2403.19708/) (2024)
- [DeepSpeed-FastGen: High-throughput Text Generation for LLMs via MII and DeepSpeed-Inference](https://mlsanity.com/paper/2401.08671/) (2024)
- [Parrot: Efficient Serving of LLM-based Applications with Semantic Variable](https://mlsanity.com/paper/2405.19888/) (2024)
- [PQCache: Product Quantization-based KVCache for Long Context LLM Inference](https://mlsanity.com/paper/2407.12820/) (2024)
- [ChunkAttention: Efficient Self-Attention with Prefix-Aware KV Cache and Two-Phase Partition](https://mlsanity.com/paper/2402.15220/) (2024)
- [Stateful Large Language Model Serving with Pensieve](https://mlsanity.com/paper/2312.05516/) (2023)
- [SagaLLM: Context Management, Validation, and Transaction Guarantees for Multi-Agent LLM Planning](https://mlsanity.com/paper/2503.11951/) (2025)
- [Autellix: An Efficient Serving Engine for LLM Agents as General Programs](https://mlsanity.com/paper/2502.13965/) (2025)
- [Continuum: Efficient and Robust Multi-Turn LLM Agent Scheduling with KV Cache Time-to-Live](https://mlsanity.com/paper/2511.02230/) (2025)
- [KVCache Cache in the Wild: Characterizing and Optimizing KVCache Cache at a Large Cloud Provider](https://mlsanity.com/paper/2506.02634/) (2025)
- [Fast State Restoration in LLM Serving with HCache](https://mlsanity.com/paper/2410.05004/) (2024)
- [Compute Or Load KV Cache? Why Not Both?](https://mlsanity.com/paper/2410.03065/) (2024)
- [Chameleon: a Heterogeneous and Disaggregated Accelerator System for Retrieval-Augmented Language Models](https://mlsanity.com/paper/2310.09949/) (2023)
- [λ-Tune: Harnessing Large Language Models for Automated Database System Tuning](https://mlsanity.com/paper/2411.03500/) (2024)
- [TokenCake: A KV-Cache-centric Serving Framework for LLM-based Multi-Agent Applications](https://mlsanity.com/paper/2510.18586/) (2025)
- [RetroInfer: A Vector Storage Engine for Scalable Long-Context LLM Inference](https://mlsanity.com/paper/2505.02922/) (2025)
- [Can Large Language Models Be Query Optimizer for Relational Databases?](https://mlsanity.com/paper/2502.05562/) (2025)
- [LEAP: LLM-powered End-to-end Automatic Library for Processing Social Science Queries on Unstructured Data](https://mlsanity.com/paper/2501.03892/) (2025)
- [Apt-Serve: Adaptive Request Scheduling on Hybrid Cache for Scalable LLM Inference Serving](https://mlsanity.com/paper/2504.07494/) (2025)
- [OrbitFlow: SLO-Aware Long-Context LLM Serving with Fine-Grained KV Cache Reconfiguration](https://mlsanity.com/paper/2601.10729/) (2026)

## Built on by (most cited first, 3 of 3)

- [SARA: SLO-Aware Resource Allocation for Disaggregated Agentic LLM Services](https://mlsanity.com/paper/2609.26763/) (2026)
- [When Fancy Eviction Fails: Rethinking Cache Replacement For LLM Prefix Reuse](https://mlsanity.com/paper/2609.28870/) (2026)
- [Accelerating Cross-Model LLM Inference via Attention Map Reuse](https://mlsanity.com/paper/2610.14688/) (2026)

## Abstract (arXiv)

Large Language Model (LLM) inference relies on key-value (KV) caches to avoid redundant attention computation. While approximate KV cache retention techniques reduce memory usage by sacrificing model accuracy, lossless approaches instead evict KV cache blocks from GPU memory and reconstruct them on demand to preserve exact outputs. Existing lossless KV cache management systems primarily base eviction decisions on access frequency or positional heuristics, without considering how different KV cache blocks affect the execution efficiency of GPU attention kernels. In this paper, we propose AsymCache, a computation-latency-aware KV cache management system for LLM inference that explicitly aligns cache residency decisions with GPU attention kernel performance, including three key components: Multi-Segment Attention (MSA) for efficient non-contiguous KV context processing, a cache eviction policy that jointly optimizes hit rate and position-aware recomputation cost, and an adaptive chunking scheduler for high hardware utilization. Experiments show that AsymCache reduces TTFT by up to 1.90-2.03x and time-per-output-token (TPOT) by 1.62-1.71x over latest baselines, confirming the effectiveness of the method in common workloads and validating its design goal of balancing computational efficiency with cache hit rate. Moreover, the low-level design of AsymCache allows seamless integration into agent serving systems such as Continuum, where it further reduces average job latency by up to 18.1%.