English
Related papers

Related papers: Nexus:Proactive Intra-GPU Disaggregation of Prefil…

200 papers

With the increased demand on economy and efficiency of measurement technology, Non-Intrusive Load Monitoring (NILM) has received more and more attention as a cost-effective way to monitor electricity and provide feedback to users. Deep…

Machine Learning · Computer Science 2020-09-28 Gan Zhou , Zhi Li , Meng Fu , Yanjun Feng , Xingyao Wang , Chengwei Huang

Deploying a large language model (LLM) inference service remains costly because centralized serving depends on specialized GPU clusters and high-bandwidth interconnects in datacenters. An appealing alternative is to leverage collaborative…

Distributed, Parallel, and Cluster Computing · Computer Science 2025-10-01 Chris Tong , Youhe Jiang , Gufeng Chen , Tianyi Zhao , Sibian Lu , Wenjie Qu , Eric Yang , Lynn Ai , Binhang Yuan

The limited HBM capacity has become the primary bottleneck for hosting an increasing number of larger-scale GPU tasks. While demand paging extends capacity via host DRAM, it incurs up to 78x slowdown due to the massive working sets and poor…

Operating Systems · Computer Science 2026-01-05 Weihang Shen , Yinqiu Chen , Rong Chen , Haibo Chen

The growing demand for large language models (LLMs) requires serving systems to handle many concurrent requests with diverse service level objectives (SLOs). This exacerbates head-of-line (HoL) blocking during the compute-intensive prefill…

Distributed, Parallel, and Cluster Computing · Computer Science 2026-02-19 Chia-chi Hsieh , Zan Zong , Xinyang Chen , Jianjiang Li , Jidong Zhai , Lijie Wen

We study offline scheduling for large language model (LLM) serving under a fixed KV-cache memory budget, where requests have heterogeneous prompt (prefill) and response (decode) lengths. Prompt tokens determine initial KV usage, and each…

Optimization and Control · Mathematics 2026-02-11 Meixuan Wang , Yinyu Ye , Zijie Zhou

Existing large language model (LLM) serving systems typically employ Prefill-Decode disaggregated architecture to prevent computational interference between the prefill and decode phases. However, in real-world LLM serving scenarios,…

Distributed, Parallel, and Cluster Computing · Computer Science 2025-11-07 Yu Wu , Tongxuan Liu , Yuting Zeng , Siyu Wu , Jun Xiong , Xianzhe Dong , Hailong Yang , Ke Zhang , Jing Li

Deploying large language models (LLMs) on mobile devices increasingly relies on heterogeneous execution, yet no prior study has systematically characterized NPU effectiveness at the operator and pipeline level. We present the first…

Hardware Architecture · Computer Science 2026-05-28 Pu Li , Jiawen Qi , Qinyu Chen

In the last decades, the computational power of GPUs has grown exponentially, allowing current deep learning (DL) applications to handle increasingly large amounts of data at a progressively higher throughput. However, network and storage…

Distributed, Parallel, and Cluster Computing · Computer Science 2025-09-08 Francesco Versaci , Giovanni Busonera

Existing LLM serving strategies can be categorized based on whether prefill and decode phases are disaggregated: non-disaggregated (NoDG) or fully disaggregated (FuDG). However, the NoDG strategy leads to strong prefill-decode interference…

Distributed, Parallel, and Cluster Computing · Computer Science 2025-04-28 Jiangsu Du , Hongbin Zhang , Taosheng Wei , Zhenyi Zheng , Kaiyi Wu , Zhiguang Chen , Yutong Lu

Performing Retrieval-Augmented Generation (RAG) directly on mobile devices is promising for data privacy and responsiveness but is hindered by the architectural constraints of mobile NPUs. Specifically, current hardware struggles with the…

Computation and Language · Computer Science 2025-12-18 Zhiyang Chen , Daliang Xu , Haiyang Shen , Chiheng Lou , Mengwei Xu , Shangguang Wang , Xin Jin , Yun Ma

Modern deployments of Large Language Models (LLMs) increasingly require serving multiple models with diverse architectures, sizes, and specialization on shared, heterogeneous hardware. This setting introduces new challenges for resource…

Artificial Intelligence · Computer Science 2026-05-20 Mert Yildiz , Pietro Spadaccino , Alexey Rolich , Francesca Cuomo , Andrea Baiocchi

Large Language Model (LLM) serving must meet stringent Service Level Objectives (SLOs) for both the prefill and decode phases. Some existing solutions disaggregate the two phases, causing potential resource idleness or compute redundancy.…

Operating Systems · Computer Science 2026-02-10 Yukang Chen , Weihao Cui , Han Zhao , Ziyi Xu , Xiaoze Fan , Xusheng Chen , Yangjie Zhou , Shixuan Sun , Bingsheng He , Quan Chen

Recent breakthroughs in Large-scale language models (LLMs) have demonstrated impressive performance on various tasks. The immense sizes of LLMs have led to very high resource demand and cost for running the models. Though the models are…

Machine Learning · Computer Science 2024-03-05 Juntao Zhao , Borui Wan , Yanghua Peng , Haibin Lin , Chuan Wu

The evolution of Large Language Models from the Transformer architecture to models with trillions of parameters has shifted the primary bottleneck from model training to real time inference. Deploying these massive models is a complex…

Distributed, Parallel, and Cluster Computing · Computer Science 2025-11-12 Madabattula Rajesh Kumar , Srinivasa Rao Aravilli , Mustafa Saify , Shashank Srivastava

Graph neural networks (GNNs) have extended the success of deep neural networks (DNNs) to non-Euclidean graph data, achieving ground-breaking performance on various tasks such as node classification and graph property prediction.…

Machine Learning · Computer Science 2021-12-17 Tianfeng Liu , Yangrui Chen , Dan Li , Chuan Wu , Yibo Zhu , Jun He , Yanghua Peng , Hongzheng Chen , Hongzhi Chen , Chuanxiong Guo

Large language models (LLMs) are increasingly deployed on customer devices. To support them, current devices are adopting SoCs (System on Chip) with NPUs (Neural Processing Unit) installed. Although high performance is expected, LLM…

Hardware Architecture · Computer Science 2025-11-17 Jianyu Wei , Qingtao Li , Shijie Cao , Lingxiao Ma , Zixu Hao , Yanyong Zhang , Xiaoyan Hu , Ting Cao

Disaggregated LLM serving improves resource efficiency by separating the compute-intensive prefill phase from the latency-critical decode phase. However, this architecture introduces a fundamental bottleneck: key/value (KV) tensors…

Distributed, Parallel, and Cluster Computing · Computer Science 2025-12-23 Dongha Yoon , Younghoon Min , Hoshik Kim , Sam H. Noh , Jongryool Kim

The rising computational and energy demands of deep learning, particularly in large-scale architectures such as foundation models and large language models (LLMs), pose significant challenges to sustainability. Traditional gradient-based…

Machine Learning · Computer Science 2025-09-19 Mohammad Saleh Vahdatpour , Huaiyuan Chu , Yanqing Zhang

Large language model (LLM) decoding is a major inference bottleneck because its low arithmetic intensity makes performance highly sensitive to memory bandwidth. 3D-stacked near-memory processing (NMP) provides substantially higher local…

Hardware Architecture · Computer Science 2026-04-10 Chenyang Ai , Yixing Zhang , Haoran Wu , Yudong Pan , Lechuan Zhao , Wenhui OU

Large language models operate in distinct compute-bound prefill followed by memory bandwidth-bound decode phases. Hybrid Mamba-Transformer models inherit this asymmetry while adding state space model (SSM) recurrences and element-wise…

Hardware Architecture · Computer Science 2026-03-17 Alish Kanani , Sangwan Lee , Han Lyu , Jiahao Lin , Jaehyun Park , Umit Y. Ogras