English
Related papers

Related papers: vAttention: Dynamic Memory Management for Serving …

200 papers

LLM decoding is bottlenecked for large batches and long contexts by loading the key-value (KV) cache from high-bandwidth memory, which inflates per-token latency, while the sequential nature of decoding limits parallelism. We analyze the…

Machine Learning · Computer Science 2025-05-28 Ted Zadouri , Hubert Strauss , Tri Dao

Large Language Models (LLMs) demonstrate substantial potential across a diverse array of domains via request serving. However, as trends continue to push for expanding context sizes, the autoregressive nature of LLMs results in highly…

Distributed, Parallel, and Cluster Computing · Computer Science 2024-07-08 Bin Lin , Chen Zhang , Tao Peng , Hanyu Zhao , Wencong Xiao , Minmin Sun , Anmin Liu , Zhipeng Zhang , Lanbo Li , Xiafei Qiu , Shen Li , Zhigang Ji , Tao Xie , Yong Li , Wei Lin

Large language model (LLM) inference demands significant amount of computation and memory, especially in the key attention mechanism. While techniques, such as quantization and acceleration algorithms, like FlashAttention, have improved…

Machine Learning · Computer Science 2024-12-18 Hao Kang , Srikant Bharadwaj , James Hensman , Tushar Krishna , Victor Ruhle , Saravan Rajmohan

Large language models (LLMs) are increasingly employed for complex tasks that process multiple generation calls in a tree structure with shared prefixes of tokens, including few-shot prompting, multi-step reasoning, speculative decoding,…

Computation and Language · Computer Science 2025-03-10 Jinwei Yao , Kaiqi Chen , Kexun Zhang , Jiaxuan You , Binhang Yuan , Zeke Wang , Tao Lin

The Transformer architecture has revolutionized deep learning, delivering the state-of-the-art performance in areas such as natural language processing, computer vision, and time series prediction. However, its core component,…

Machine Learning · Computer Science 2025-07-24 Haolin Jin , Mengbai Xiao , Yuan Yuan , Xiao Zhang , Dongxiao Yu , Guanghui Zhang , Haoliang Wang

Large Language Model (LLM) inference is increasingly constrained by memory bandwidth, with frequent access to the key-value (KV) cache dominating data movement. While attention sparsity reduces some memory traffic, the relevance of past…

Hardware Architecture · Computer Science 2025-09-16 Yunhua Fang , Rui Xie , Asad Ul Haq , Linsen Ma , Kaoutar El Maghraoui , Naigang Wang , Meng Wang , Liu Liu , Tong Zhang

Large Language Models (LLMs) exhibit pronounced memory-bound characteristics during inference due to High Bandwidth Memory (HBM) bandwidth constraints. In this paper, we propose an L2 Cache-oriented asynchronous KV Cache prefetching method…

Machine Learning · Computer Science 2025-11-11 Yanhao Dong , Yubo Miao , Weinan Li , Xiao Zheng , Chao Wang , Jiesheng Wu , Feng Lyu

Multimodal large language models (MLLMs) are plagued by exorbitant inference costs attributable to the profusion of visual tokens within the vision encoder. The redundant visual tokens engenders a substantial computational load and…

Computer Vision and Pattern Recognition · Computer Science 2026-02-03 Jiedong Zhuang , Lu Lu , Ming Dai , Rui Hu , Jian Chen , Qiang Liu , Haoji Hu

In this study, we introduce adaptive KV cache compression, a plug-and-play method that reduces the memory footprint of generative inference for Large Language Models (LLMs). Different from the conventional KV cache that retains key and…

Computation and Language · Computer Science 2024-10-31 Suyu Ge , Yunan Zhang , Liyuan Liu , Minjia Zhang , Jiawei Han , Jianfeng Gao

The memory and computational demands of Key-Value (KV) cache present significant challenges for deploying long-context language models. Previous approaches attempt to mitigate this issue by selectively dropping tokens, which irreversibly…

Machine Learning · Computer Science 2024-07-24 Hanlin Tang , Yang Lin , Jing Lin , Qingsen Han , Shikuan Hong , Yiwu Yao , Gongyi Wang

Transformer-based Large Language Models (LLMs) have become increasingly important. However, due to the quadratic time complexity of attention computation, scaling LLMs to longer contexts incurs extremely slow inference speed and high GPU…

Attention accounts for an increasingly dominant fraction of total computation during inference for mixture-of-experts (MoE) models, making efficient acceleration critical. Emerging domain-specific accelerators for large model inference are…

Hardware Architecture · Computer Science 2026-04-03 Chi Zhang , Luca Colagrande , Renzo Andri , Luca Benini

System-level routers that intercept LLM requests for safety classification, domain routing, and PII detection must be both fast and operationally lightweight: they should add minimal latency to every request, yet not require a dedicated GPU…

Computation and Language · Computer Science 2026-03-16 Xunzhuo Liu , Bowei He , Xue Liu , Andy Luo , Haichen Zhang , Huamin Chen

KV cache quantization reduces the memory cost of long-context LLM inference, but introduces approximation error that is typically validated only empirically. Existing systems rely on average-case robustness, with no mechanism to detect or…

Machine Learning · Computer Science 2026-05-21 Dean Calver

LLMs are seeing growing use for applications which require large context windows, and with these large context windows KV cache activations surface as the dominant contributor to memory consumption during inference. Quantization is a…

Machine Learning · Computer Science 2025-05-30 Coleman Hooper , Sehoon Kim , Hiva Mohammadzadeh , Michael W. Mahoney , Yakun Sophia Shao , Kurt Keutzer , Amir Gholami

FlashAttention (Dao, 2023) effectively reduces the quadratic peak memory usage to linear in training transformer-based large language models (LLMs) on a single GPU. In this paper, we introduce DISTFLASHATTN, a distributed memory-efficient…

Machine Learning · Computer Science 2024-04-02 Dacheng Li , Rulin Shao , Anze Xie , Eric P. Xing , Xuezhe Ma , Ion Stoica , Joseph E. Gonzalez , Hao Zhang

Key-Value (KV) cache plays a crucial role in accelerating inference in large language models (LLMs) by storing intermediate attention states and avoiding redundant computation during autoregressive generation. However, its memory footprint…

Machine Learning · Computer Science 2026-04-14 Yuzhen Mao , Qitong Wang , Martin Ester , Ke Li

How to efficiently serve LLMs in practice has become exceptionally challenging due to their prohibitive memory and computation requirements. In this study, we investigate optimizing the KV cache, whose memory footprint poses a critical…

Computation and Language · Computer Science 2025-06-10 Akshat Sharma , Hangliang Ding , Jianping Li , Neel Dani , Minjia Zhang

TiledAttention is a scaled dot-product attention (SDPA) forward operator for SDPA research on NVIDIA GPUs. Implemented in cuTile Python (TileIR) and exposed as a PyTorch-callable function, it is easier to modify than low-level CUDA…

Machine Learning · Computer Science 2026-05-12 Taimur Khan

Large Language Model (LLM) serving is increasingly constrained by the growing size of the key-value (KV) cache, which scales with both context length and generation length. Prior work shows that attention is dominated by a small subset of…

Machine Learning · Computer Science 2026-04-21 Nazmul Takbir , Hamidreza Alikhani , Nikil Dutt , Sangeetha Abdu Jyothi