arXiv AI

Elastic KV Cache for LLM Serving:A Working Reclamation Mechanism, and Why Chunked Prefill Already Closes the Gap

The paper introduces an elastic key‑value (KV) cache for large language model (LLM) serving that dynamically reclaims a pre‑allocated reserve during decode‑heavy phases and restores it before prefill, using a userspace CUDA virtual‑memory trick that requires no driver changes. The authors implement this mechanism, test it under realistic workloads, and find that it offers only marginal benefits—about a 1 % difference in time‑to‑first‑token for large prefill chunks—and that simpler strategies such as lowering the maximum batch size can achieve similar results. The study also notes that the reserve’s impact diminishes with higher tensor‑parallelism levels. whyItMatters":"The work demonstrates that a dynamic KV cache reclamation strategy can be implemented without driver patches and that its practical benefits are limited, guiding future LLM serving optimizations toward simpler approaches."

arXiv Machine Learning

Sep 11

Building py-kvcache: A Performance Characterization of External KV Caching for vLLM with NVMe SSDs

The paper presents py‑kvcache, a new KV offload connector for vLLM that uses asynchronous direct I/O, bounded shared staging, and scheduler‑aware preloading to improve external KV caching performance on NVMe SSDs. Experiments across synthetic workloads, long‑context benchmarks, and production traces show that py‑kvcache can load 80k‑token prefixes 2.0× faster than LMCache, with preloading contributing an additional 1.34× speedup, and achieves overall performance within 4% of native vLLM KV Offload. The study highlights that cache effectiveness depends on transfer granularity, intermediate memory use, and scheduling timing rather than just device bandwidth, indicating that external KV caching should be considered a setup‑specific admission decision.

By Joseph Kanichai, Tiziano De Matteis, Animesh Trivedi

llms benchmarks

arXiv AI

Jul 3

Towards Load-Aware Prefill Deflection for Disaggregated LLM Serving

arXiv:2607. 02043v1 Announce Type: cross Abstract: Disaggregated LLM serving runs prefill and decode on separate GPU pools to keep the two phases from interfering.

By Shrikara Arun, Anjaly Parayil, Srikant Bharadwaj, Renee St. Amant, Victor R\"uhle

llms benchmarks

Hugging Face Trending Papers

Jul 5

Sangam: Efficiently Serving Diffusion LLMs with the AR Stack

Diffusion language models (dLLMs) generate text by iteratively denoising a masked response and can commit multiple output positions per model invocation. Their bidirectional attention prevents exact autoregressive-style KV caching, since committing one position shifts the KV activations of all others.

llms diffusion

arXiv AI

Jul 1

KV-RM: Regularizing KV-Cache Movement for Static-Graph LLM Serving

arXiv:2605. 09735v2 Announce Type: replace-cross Abstract: Static-graph LLM decoders provide predictable launches, fixed tensor shapes, and low submission overhead, but online decoding exposes highly irregular KV-cache behavior: request lengths differ, EOS events arrive asynchronously, and logical histories fragment over time.

By Zhiqing Zhong, Zhijing Ye, Jian Zhang, Weijian Zheng, Bolun Sun, Xiaodong Yu

llms efficiency

arXiv Machine Learning

3d ago

Physically Partitioned KVCache Format for CPU--GPU Load Balancing in MoE Inference

arXiv:2609.14507v1 Announce Type: cross Abstract: Single-GPU long-context inference with Mixture-of-Experts (MoE) models requires spilling the key-value cache (KVCache) to CPU memory. The spilled KV...

By Enda Yu, Dezun Dong, Xiangke Liao

arXiv AI

Jun 24

CrossPool: Efficient Multi-LLM Serving for Cold MoE Models through KV-Cache and Weight Disaggregation

arXiv:2606. 24506v1 Announce Type: cross Abstract: Emerging LLM services increasingly host many sparse MoE models, yet most models receive sparse requests and remain cold.

By Zhuoren Ye, Tianyu Wo, Dinghao Xue, Mingming Zhang, Yuchen Teng, Chunming Hu, Renyu Yang

llms benchmarks

arXiv AI

Sep 10

Characterizing Contention-Induced Reliability Collapse in KV-Cache Timing Side Channels for Multi-Tenant LLM Serving

arXiv:2609.06853v1 Announce Type: cross Abstract: Shared key--value (KV) cache reuse improves large language model (LLM) serving, but it can also create a timing side channel that reveals whether a p...

By Rana Abu Bakar

llms

arXiv Machine Learning

Jun 5

Tangram: Unlocking Non-Uniform KV Cache for Efficient Multi-turn LLM Serving

arXiv:2606. 06302v1 Announce Type: new Abstract: Multi-turn Large Language Model (LLM) serving is critical for consistent user experiences, yet the linear growth of the Key-Value (KV) cache imposes significant pressure on GPU memory and bandwidth.

By Hyungmin Kim, Minsoo Kim, Hongseok Kim, Jungwook Choi

llms efficiency

arXiv AI

Aug 14

vToken: Token-Level Virtualization for Reclaimable KV Caches

arXiv:2608. 13263v1 Announce Type: new Abstract: Large language model serving faces a critical memory bottleneck: the KV cache grows with sequence length and batch size.

By Yuanhang Gao, Xiangrui Yang, Yuanfeng Chen, Hongjia Chen, Qianru Lv, Wenfei Wu, Dongsheng Li

llms efficiency

arXiv Machine Learning

3d ago

Shared KV Caching for Replicated 27B Inference: Correctness Failures and Performance Boundaries

arXiv:2609.15021v1 Announce Type: cross Abstract: Shared host-memory caching can avoid repeated prefill when a request moves between inference replicas. Its usefulness depends on both correct state t...

By Frank Li

arXiv Machine Learning

Jul 30

InferScale: GPU-Native KV Injection for Personalized LLM Serving

arXiv:2607. 27090v1 Announce Type: cross Abstract: Large language models are increasingly deployed with persistent personalized context, such as accumulated memory profiles or long conversation histories, that is shared across a user's many requests.

By Peter Li, Prashant Pandey

llms rag fine-tuning efficiency

arXiv Machine Learning

Sep 1

WiSP: A Working-Set View of Mixture-of-Experts Serving on Extremely Low-Resource Hardware

WiSP (Working‑Set Paging) is a routing‑aware expert pager that allows Mixture‑of‑Experts models to run on GPUs that cannot hold the entire expert pool by paging experts in and out of VRAM while preserving byte‑identical outputs. On a 24 GiB RTX 3090, WiSP doubles decode throughput compared to static offload when the model does not fit, and its companion policy MV‑WSA allocates VRAM between resident experts and KV cache based on marginal latency benefit, reducing end‑to‑end time by up to 1.19× without altering model outputs.

By Jiamu Zhang, Liang Wu, Mayank Darbari, Liangjie Hong

agents efficiency

Related stories