Searcharxiv⌕ Search

arXiv subjects

Kuntai Du

Publications and source records attributed to Kuntai Du.

At least 19 recordsLinked to original sources
arXiv2026

Learning Agent Execution for KV-Cache Management in Agentic Serving

Multi-agent LLM systems have emerged as an important deployment paradigm for AI services, where each user request is decomposed into a sequence of specialized agents. Across these workflows, every agent repeatedly executes a fixed context consisting of system prompts, tool definitions, and few-shot examples, creating substantial opportunities for KV-cache reuse. Existing LLM serving systems, however, manage KV-cache reactively using prefix caching and recency-based replacement, causing reusable agent contexts to be evicted before their next invocation and forcing repeated recomputation. We present CacheScout, an agent-aware KV-cache runtime layer for multi-agent LLM serving. The key insight is that future KV-cache reuse is governed by agent execution semantics rather than cache recency alone. CacheScout captures these semantics by learning agent execution transitions online, without requiring predefined workflow graphs or offline training, and uses the learned execution model to guide both cache eviction and proactive prefetching while leaving the serving critical path unchanged. We implement CacheScout on top of vLLM. Across representative real-world multi-agent workloads, CacheScout improves KV-cache hit rate by 10-18 percentage points, reduces mean TTFT by 18-45%, lowers mean per-turn latency by 29-38%, and increases peak throughput by up to 57%. These benefits also generalize to larger models, reducing TTFT by up to 54% while sustaining 37% higher throughput.

cs.AI↗
arXiv2026

VeriCache: Turning Lossy KV Cache into Lossless LLM Inference

The large size of the KV cache has become a major bottleneck for serving LLMs with increasing context lengths. In response, many KV cache compression methods, such as token dropping and quantization, have been proposed. However, almost all of these methods are inherently lossy-despite minimal accuracy degradation for short outputs, their outputs increasingly diverge from full-KV-cache outputs as more tokens are decoded, which leads to catastrophic failures in code generation and tool calling. We present VeriCache, the first inference framework that ensures the same output as full-KV-cache decoding but largely preserves the high decoding throughput of a range of KV cache compression algorithms. VeriCache uses the compressed KV cache to draft tokens, then verifies them against the full KV cache. While it may seem like just speculative decoding, VeriCache requires addressing a key system challenge to work-keeping the full KV cache out of GPU memory and minimizing the overhead of swapping it in for verification. The insight is two-fold: (1) compressed-KV decoding can be parallelized with full-KV swap, because one is HBM-bandwidth-bound and the other is PCIe/network-bound, and (2) the compressed KV cache often produces output similar to the full KV cache, allowing a long drafting horizon to amortize each full-KV swap. VeriCache applies to both long-context decoding and remote prefix caching, supports a broad family of token-dropping and quantization methods through a uniform compressor interface, and composes with traditional speculative decoding. Experimental results show that VeriCache achieves up to 4X higher throughput than full-KV inference while producing identical outputs.

cs.AR↗
arXiv2026

AdaptCache: KV Cache Native Storage Hierarchy for Low-Delay and High-Quality Language Model Serving

Large language model (LLM) applications often reuse previously processed context, such as chat history and documents, which introduces significant redundant computation. Existing LLM serving systems address such redundant computation by storing the KV caches of processed context and loading the corresponding KV cache when a new request reuses the context. Further, as these LLM applications scale, the total size of KV caches becomes excessively large and requires both DRAM and SSD for full storage. However, prior work that stores KV caches in DRAM and SSD suffers from high loading delays, as most KV cache hits come from SSD, which is slow to load. To increase the KV cache hit rate on DRAM, we identify lossy KV cache compression as a promising approach. We design a lossy compression system that decides the compression algorithm, compression rate and device placement for each KV cache entry to maximise DRAM hits and minimise loading delay without significantly degrading generation quality. Compared to various static compression baselines across three tasks, our system AdaptCache achieves 1.43--2.4 x delay savings at the same quality and 6--55% quality improvements at the same delay.

cs.OS↗
arXiv2025

EVICPRESS: Joint KV-Cache Compression and Eviction for Efficient LLM Serving

Reusing KV cache is essential for high efficiency of Large Language Model (LLM) inference systems. With more LLM users, the KV cache footprint can easily exceed GPU memory capacity, so prior work has proposed to either evict KV cache to lower-tier storage devices, or compress KV cache so that more KV cache can be fit in the fast memory. However, prior work misses an important opportunity: jointly optimizing the eviction and compression decisions across all KV caches to minimize average generation latency without hurting quality. We propose EVICPRESS, a KV-cache management system that applies lossy compression and adaptive eviction to KV cache across multiple storage tiers. Specifically, for each KV cache of a context, EVICPRESS considers the effect of compression and eviction of the KV cache on the average generation quality and delay across all contexts as a whole. To achieve this, EVICPRESS proposes a unified utility function that quantifies the effect of quality and delay of the lossy compression or eviction. To this end, EVICPRESS's profiling module periodically updates the utility function scores on all possible eviction-compression configurations for all contexts and places KV caches using a fast heuristic to rearrange KV caches on all storage tiers, with the goal of maximizing the utility function scores on each storage tier. Compared to the baselines that evict KV cache or compress KV cache, EVICPRESS achieves higher KV-cache hit rates on fast devices, i.e., lower delay, while preserving high generation quality by applying conservative compression to contexts that are sensitive to compression errors. Evaluation on 12 datasets and 5 models demonstrates that EVICPRESS achieves up to 2.19x faster time-to-first-token (TTFT) at equivalent generation quality.

cs.OS↗
arXiv2025

LMCache: An Efficient KV Cache Layer for Enterprise-Scale LLM Inference

KV cache has traditionally been stored in GPU memory to accelerate the decoding phase of large language model (LLM) inference. However, it is increasingly necessary to move KV caches outside GPU devices, to enable cache reuse across different queries and inference engines. Our real-world usage statistics confirm this trend: over time, the total KV cache stored by users has grown rapidly, far exceeding the capacity of GPU memory. Despite this need, there lacks an efficient solution for offloading and transferring KV caches. We present LMCACHE, the first and so far the most efficient open-source KV caching solution, which extracts and stores KV caches generated by modern LLM engines (vLLM and SGLang) out of the GPU memory and shares them across engines and queries. LMCACHE supports both cache offloading (prefix reuse across queries) and prefill-decode (PD) disaggregation (cross-engine/GPU cache transfer). LMCACHE's high performance and wide adoption stem from the following contributions: (1) highly optimized KV cache data movement powered by batched data movement operations, compute and I/O pipelining; (2) a modular KV cache connector component, decoupling LMCACHE from the rapid evolution of inference engines; (3) a first-class control API for flexible cache orchestration across GPU, CPU, storage, and network layers. Our evaluation shows that combining LMCACHE with vLLM achieves up to 15x improvement in throughput across workloads such as multi-round question answering and document analysis. Large-scale adoption of LMCACHE in enterprise settings provides us valuable insights, for example, fetching KV cache from remote storage has unsurprisingly benefits to prefill delay, and that context truncation, which is a widely applied technique in industry, can greatly reduce prefix cache hit ratio by half. The source code of LMCACHE is at: https://github.com/LMCache/LMCache.

cs.LG↗
arXiv2025

METIS: Fast Quality-Aware RAG Systems with Configuration Adaptation

RAG (Retrieval Augmented Generation) allows LLMs (large language models) to generate better responses with external knowledge, but using more external knowledge often improves generation quality at the expense of response delay. Prior work either reduces the response delay (through better scheduling of RAG queries) or strives to maximize quality (which involves tuning the RAG workflow), but they fall short in optimizing the tradeoff between the delay and quality of RAG responses. This paper presents METIS, the first RAG system that jointly schedules queries and adapts the key RAG configurations of each query, such as the number of retrieved text chunks and synthesis methods, in order to balance quality optimization and response delay reduction. Using 4 popular RAG-QA datasets, we show that compared with the state-of-the-art RAG optimization schemes, METIS reduces the generation latency by $1.64-2.54\times$ without sacrificing generation quality.

cs.LG↗
arXiv2025

TurboSpec: Closed-loop Speculation Control System for Optimizing LLM Serving Goodput

Large Language Model (LLM) serving systems batch concurrent user requests to achieve efficient serving. However, in real-world deployments, such inter-request parallelism from batching is often limited by external factors such as low request rates or memory constraints. Recent works focus on intra-request parallelism from speculative decoding as a solution to this problem. Unfortunately, benefits from intra-request parallelism are often fragile, as speculative decoding causes overhead, and speculated tokens may miss. We observe that speculative decoding may degrade LLM serving performance if added naively without tuning to the incoming requests and the speculation method. To alleviate the need for expert tuning and make speculative decoding more robust, we present TurboSpec, a speculation control system that automatically profiles the execution environment and utilizes a feedback-based algorithm to dynamically adjust the amount of intra-request parallelism in LLM serving. TurboSpec predicts "goodput" - the amount of successfully generated tokens - to evaluate and adjust intra-request parallelism amount to that with the highest goodput in runtime. We implement TurboSpec on a real-world LLM serving system vLLM and demonstrate its effectiveness across diverse workloads and hardware configurations, providing consistent performance improvements across all test scenarios.

cs.AI↗
arXiv2025

DroidSpeak: KV Cache Sharing for Cross-LLM Communication and Multi-LLM Serving

Compound AI systems, such as agentic systems, are an emerging trend in large-scale enterprise settings, with multiple LLMs specialized for different users, tasks, and/or roles working together. In these scenarios, different models often process inputs that share the same context prefix. Although much work was done in the past to enable the reuse of prefix KV caches across inputs for a single model, how to enable one model to reuse the prefix KV caches of a different model remains an open question. We introduce DroidSpeak, the first distributed LLM inference system that enables KV cache reuse across distributed nodes running inference of different LLMs, so long as the LLMs have the same architecture. We present the first study that aims at understanding the impact of sharing KV caches across different LLMs, and if/when such sharing affects quality. Inspired by the findings, we present DroidSpeak, which selectively recomputes a few layers of the KV cache produced by another LLM and reuses the remaining layers, with negligible quality loss. Moreover, carefully pipelining the layer-wise re-computation and the loading of reused KV cache further improves the inference performance. Experiments on diverse datasets and model pairs demonstrate that DroidSpeak achieves up to 4x throughput improvement and about 3.1x faster prefill (time to first token), with negligible loss of quality in F1 scores, Rouge-L or code similarity score, compared to the baseline which does not allow any sharing across models.

cs.MA↗
arXiv2025

PrefillOnly: An Inference Engine for Prefill-only Workloads in Large Language Model Applications

Besides typical generative applications, like ChatGPT, GitHub Copilot, and Cursor, we observe an emerging trend that LLMs are increasingly used in traditional discriminative tasks, such as recommendation, credit verification, and data labeling. The key characteristic of these emerging use cases is that the LLM generates only a single output token, rather than an arbitrarily long sequence of tokens. We call this prefill-only workload. However, since existing LLM engines assume arbitrary output lengths, they fail to leverage the unique properties of prefill-only workloads. In this paper, we present PrefillOnly, the first LLM inference engine that improves the inference throughput and latency by fully embracing the properties of prefill-only workloads. First, since it generates only one token, PrefillOnly only needs to store the KV cache of only the last computed layer, rather than of all layers. This drastically reduces the GPU memory footprint of LLM inference and allows handling long inputs without using solutions that reduces throughput, such as cross-GPU KV cache parallelization. Second, because the output length is fixed, rather than arbitrary, PrefillOnly can precisely determine the job completion time (JCT) of each prefill-only request before it starts. This enables efficient JCT-aware scheduling policies such as shortest remaining job first. PrefillOnly can process upto 4x larger queries per second without inflating average and P99 latency.

cs.DC↗
arXiv2025

CacheBlend: Fast Large Language Model Serving for RAG with Cached Knowledge Fusion

Large language models (LLMs) often incorporate multiple text chunks in their inputs to provide the necessary contexts. To speed up the prefill of the long LLM inputs, one can pre-compute the KV cache of a text and re-use the KV cache when the context is reused as the prefix of another LLM input. However, the reused text chunks are not always the input prefix, which makes precomputed KV caches not directly usable since they ignore the text's cross-attention with the preceding texts. Thus, the benefits of reusing KV caches remain largely unrealized. This paper tackles just one challenge: when an LLM input contains multiple text chunks, how to quickly combine their precomputed KV caches in order to achieve the same generation quality as the expensive full prefill (i.e., without reusing KV cache)? This challenge naturally arises in retrieval-augmented generation (RAG) where the input is supplemented with multiple retrieved texts as the context. We present CacheBlend, a scheme that reuses the precomputed KV caches, regardless prefix or not, and selectively recomputes the KV values of a small subset of tokens to partially update each reused KV cache. In the meantime, the small extra delay for recomputing some tokens can be pipelined with the retrieval of KV caches within the same job, allowing CacheBlend to store KV caches in slower devices with more storage capacity while retrieving them without increasing the inference delay. By comparing CacheBlend with the state-of-the-art KV cache reusing schemes on three open-source LLMs of various sizes and four popular benchmark datasets of different tasks, we show that CacheBlend reduces time-to-first-token (TTFT) by 2.2-3.3x and increases the inference throughput by 2.8-5x from full KV recompute without compromising generation quality. The code is available at https://github.com/LMCache/LMCache.

cs.LG↗
arXiv2025

Jenga: Effective Memory Management for Serving LLM with Heterogeneity

Large language models (LLMs) are widely used but expensive to run, especially as inference workloads grow. To lower costs, maximizing the request batch size by managing GPU memory efficiently is crucial. While PagedAttention has recently been proposed to improve the efficiency of memory management, we find that the growing heterogeneity in the embeddings dimensions, attention, and access patterns of modern LLM architectures introduces new challenges for memory allocation. In this paper, we present Jenga, a novel memory allocation framework for heterogeneous embeddings in LLMs. Jenga tackles two key challenges: (1) minimizing memory fragmentation when managing embeddings of different sizes, and (2) enabling flexible caching and eviction policies tailored to the specific token-dependency patterns of various layers. Jenga employs a two-level memory allocator, leveraging the least common multiple (LCM) of embedding sizes to optimize memory usage and providing APIs to express layer-specific caching logic to enhance memory reuse. We implemente Jenga on vLLM, a state-of-the-art LLM inference engine, and evaluate it with diverse LLMs, datasets, and GPU configurations. Evaluations show that Jenga improves GPU memory utilization by up to 79.6%, and increases serving throughput by up to 4.92x (1.80x on average).

cs.DC↗
arXiv2025

Towards More Economical Context-Augmented LLM Generation by Reusing Stored KV Cache

Across large language model (LLM) applications, we observe an emerging trend for reusing KV caches to save the prefill delays of processing repeated input texts in different LLM inputs. This has led to a broad design space, including colocating stored KV caches with (or close to) GPUs to various KV cache compression. However, a key question remains unanswered: can these delay reductions also be economically favorable? Specifically, we ask whether a developer can use public cloud services to store precomputed KV caches and reuse them to save delay without incurring more costs in terms of compute, storage, and network. To answer this question, we propose an validated analytical model for the cloud cost (in compute, storage, and network) of storing and reusing KV caches based on various workload parameters, such as reuse frequency, generated text lengths, model sizes, etc. Preliminary results show that KV cache reusing is able to save both delay and cloud cost across a range of workloads with long context. And we call more efforts on building more economical context augmented LLM by KV cache reusing.

cs.NI↗
arXiv2024

OneAdapt: Fast Configuration Adaptation for Video Analytics Applications via Backpropagation

Deep learning inference on streaming media data, such as object detection in video or LiDAR feeds and text extraction from audio waves, is now ubiquitous. To achieve high inference accuracy, these applications typically require significant network bandwidth to gather high-fidelity data and extensive GPU resources to run deep neural networks (DNNs). While the high demand for network bandwidth and GPU resources could be substantially reduced by optimally adapting the configuration knobs, such as video resolution and frame rate, current adaptation techniques fail to meet three requirements simultaneously: adapt configurations (i) with minimum extra GPU or bandwidth overhead; (ii) to reach near-optimal decisions based on how the data affects the final DNN's accuracy, and (iii) do so for a range of configuration knobs. This paper presents OneAdapt, which meets these requirements by leveraging a gradient-ascent strategy to adapt configuration knobs. The key idea is to embrace DNNs' differentiability to quickly estimate the accuracy's gradient to each configuration knob, called AccGrad. Specifically, OneAdapt estimates AccGrad by multiplying two gradients: InputGrad (i.e. how each configuration knob affects the input to the DNN) and DNNGrad (i.e. how the DNN input affects the DNN inference output). We evaluate OneAdapt across five types of configurations, four analytic tasks, and five types of input data. Compared to state-of-the-art adaptation schemes, OneAdapt cuts bandwidth usage and GPU usage by 15-59% while maintaining comparable accuracy or improves accuracy by 1-5% while using equal or fewer resources.

cs.LG↗
arXiv2024

LLMSteer: Improving Long-Context LLM Inference by Steering Attention on Reused Contexts

As large language models (LLMs) show impressive performance on complex tasks, they still struggle with longer contextual understanding and high computational costs. To balance efficiency and quality, we introduce LLMSteer, a fine-tuning-free framework that enhances LLMs through query-independent attention steering. Tested on popular LLMs and datasets, LLMSteer narrows the performance gap with baselines by 65.9% and reduces the runtime delay by up to 4.8x compared to recent attention steering methods.

cs.LG↗
arXiv2024

Do Large Language Models Need a Content Delivery Network?

As the use of large language models (LLMs) expands rapidly, so does the range of knowledge needed to supplement various LLM queries. Thus, enabling flexible and efficient injection of new knowledge in LLM inference is critical. Three high-level options exist: (i) embedding the knowledge in LLM's weights (i.e., fine-tuning), (ii) including the knowledge as a part of LLM's text input (i.e., in-context learning), or (iii) injecting the KV caches of the new knowledge to LLM during prefill. This paper argues that, although fine-tuning and in-context learning are popular, using KV caches as the medium of knowledge could simultaneously enable more modular management of knowledge injection and more efficient LLM serving with low cost and fast response. To realize these benefits, we envision a Knowledge Delivery Network (KDN), a new system component in LLM services that dynamically optimizes the storage, transfer, and composition of KV cache across LLM engines and other compute and storage resources. We believe that, just like content delivery networks (CDNs), such as Akamai, enabled the success of the Internet ecosystem through their efficient data delivery, KDNs will be critical to the success of LLM applications through their efficient knowledge delivery. We have open-sourced a KDN prototype at https://github.com/LMCache/LMCache.

cs.CL↗
arXiv2024

CacheGen: KV Cache Compression and Streaming for Fast Large Language Model Serving

As large language models (LLMs) take on complex tasks, their inputs are supplemented with longer contexts that incorporate domain knowledge. Yet using long contexts is challenging, as nothing can be generated until the whole context is processed by the LLM. While the context-processing delay can be reduced by reusing the KV cache of a context across different inputs, fetching the KV cache, which contains large tensors, over the network can cause high extra network delays. CacheGen is a fast context-loading module for LLM systems. First, CacheGen uses a custom tensor encoder, leveraging KV cache's distributional properties to encode a KV cache into more compact bitstream representations with negligible decoding overhead, to save bandwidth usage. Second, CacheGen adapts the compression level of different parts of a KV cache to cope with changes in available bandwidth, in order to maintain low context-loading delay and high generation quality. % When available bandwidth drops, CacheGen may raise the compression level for a part of the context or recompute its KV cache on the fly. We test CacheGen on popular LLMs and datasets. Compared to the recent systems that reuse the KV cache, CacheGen reduces the KV cache size by 3.5-4.3x and the total delay in fetching and processing contexts by 3.2-3.7x with negligible impact on the LLM response quality. Our code is at: https://github.com/UChi-JCL/CacheGen.

cs.NI↗
arXiv2024

Eloquent: A More Robust Transmission Scheme for LLM Token Streaming

To render each generated token in real-time for users, the Large Language Model (LLM) server generates tokens one by one and streams each token (or group of a few tokens) through the network to the user right after generation, which we refer to as LLM token streaming. However, under unstable network conditions, the LLM token streaming experience could suffer greatly from stalls since one packet loss could block the rendering of later tokens even if the packets containing them arrive on time. With a measurement study, we show that current applications suffer from increased stalls under unstable networks. For this emerging token streaming problem in LLM Chatbots that differs from previous multimedia and text applications, we propose a novel transmission scheme, called Eloquent, which puts newly generated tokens as well as currently unacknowledged tokens in the next outgoing packet. This ensures that each packet contains some new tokens and, in the meantime, is independently rendered when received, avoiding the aforementioned stalls caused by missing packets. Through simulation under various networks, we show Eloquent reduces stall ratio (proportion of token rendering wait time) by 71.0% compared to the retransmission method commonly used by real chatbot applications and by 31.6% compared to the baseline packet duplication scheme. By tailoring Eloquent to fit the token-by-token generation of LLM, we enable the Chatbots to respond like an eloquent speaker for users to better enjoy pervasive AI.

cs.NI↗
arXiv2024

Earth+: on-board satellite imagery compression leveraging historical earth observations

With the increasing deployment of earth observation satellite constellations, the downlink (satellite-to-ground) capacity often limits the freshness, quality, and coverage of the imagery data available to applications on the ground. To overcome the downlink limitation, we present Earth+, a new satellite imagery compression system that, instead of compressing each image individually, pinpoints and downloads only recent imagery changes with respect to the history reference images. To minimize the amount of changes, it is critical to make reference images as fresh as possible. Earth+ enables each satellite to choose fresh reference images from not only its own history images but also past images of other satellites from an entire satellite constellation. To share reference images across satellites, Earth+ utilizes the limited capacity of the existing uplink (ground-to-satellite) by judiciously selecting and compressing reference images while still allowing accurate change detection. In short, Earth+ is the first to make reference-based compression efficient, by enabling constellation-wide sharing of fresh reference images across satellites. Our evaluation shows that Earth+ can reduce the downlink usage by a factor of 3.3 compared to state-of-the-art on-board image compression techniques while not sacrificing image quality, or using more on-board computing or storage resources, or more uplink bandwidth than currently available.

cs.NI↗