paper-with-me

홈 › Papers

Accelerating Local LLMs on Resource-Constrained Edge Devices via Distributed Prompt Caching

2026-02-26 · Hiroki Matsutani, Naoki Matsuda, Naoto Sugiura arxiv

Since local LLM inference on resource-constrained edge devices imposes a severe performance bottleneck, this paper proposes distributed prompt caching to enhance inference performance by cooperatively sharing intermediate processing states across multiple low-end edge devices. To fully utilize prompt similarity, our distributed caching mechanism also supports partial matching. As this approach introduces communication overhead associated with state sharing over a wireless network, we introduce a Bloom-filter-based data structure, referred to as a catalog, to determine whether a remote server possesses the desired internal states, thereby suppressing unnecessary communication. Experiments using the Gemma-3 270M model and the MMLU dataset on the Raspberry Pi Zero 2W platform demonstrate that the proposed approach reduces TTFT (Time to First Token) and TTLT (Time to Last Token) by 93.12% and 50.07% on average, respectively.

📄 PDF Abstract BibTeX arXiv:2602.22812

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

FastTTS: Accelerating Test-Time Scaling for Edge LLM Reasoning

2025-08-29 · Hao Mark Chen, Zhiwen Mo, Guanxi Lu, Shuang Liang 외 arxiv

Recent advances in reasoning Large Language Models (LLMs) are driving the emergence of agentic AI systems. Edge deployment of LLM agents near end users is increasingly necessary to protect data privacy, enable offline us…

lm-Meter: Unveiling Runtime Inference Latency for On-Device Language Models

2025-10-07 · Haoxin Wang, Xiaolong Tu, Hongyu Ke, Huirong Chai 외 arxiv

Large Language Models (LLMs) are increasingly integrated into everyday applications, but their prevalent cloud-based deployment raises growing concerns around data privacy and long-term sustainability. Running LLMs local…

Accelerating Large Language Model Inference with Self-Supervised Early Exits

2024-07-30 · Florian Valade

This paper presents a novel technique for accelerating inference in large, pre-trained language models (LLMs) by introducing early exits during inference. The computational demands of these models, used across a wide ran…

Language ModelingLanguage ModellingLarge Language Model

DeepFusion: Accelerating MoE Training via Federated Knowledge Distillation from Heterogeneous Edge Devices

2026-02-15 · Songyuan Li, Jia Hu, Ahmed M. Abdelmoniem, Geyong Min 외 arxiv

Recent Mixture-of-Experts (MoE)-based large language models (LLMs) such as Qwen-MoE and DeepSeek-MoE are transforming generative AI in natural language processing. However, these models require vast and diverse training …

Knowledge DistillationFederated Learning

On Accelerating Edge AI: Optimizing Resource-Constrained Environments

2025-01-25 · Jacob Sander, Achraf Cohen, Venkat R. Dasari, Brent Venable 외

Resource-constrained edge deployments demand AI solutions that balance high performance with stringent compute, memory, and energy limitations. In this survey, we present a comprehensive overview of the primary strategie…

Knowledge DistillationModel CompressionNeural Architecture SearchQuantization+2