Skip to content
Hot storiesContinuously updated

研究团队提出移动端稀疏Prefill加速框架TileWeave

1 reports1 reporting sourcesUpdated 20 hours ago

Start here

AI overview

2026年11月3日,arXiv机器学习领域发布论文,提出了一种名为TileWeave的面向移动端SoC的异构分块方案。该技术旨在加速大语言模型的长上下文稀疏预填充(Prefill)过程,其核心方法是采用异构协同策略,将具有高复用特性的key块分流至NPU进行计算,而低复用块则交由GPU执行,以此提升端侧长文本预填充的处理效率。

AI-generated from reports · Updated 16 hours ago

Report timeline

Follow the reports to see different sides of the story.

Nov 3
  1. arXiv 机器学习
    TileWeave:面向移动端 SoC 稀疏预填充的异构分块技术

    TileWeave 提出面向移动 SoC 的异构分块方案,将高复用 key 块分流至 NPU、低复用块交由 GPU 执行,以加速大语言模型长上下文稀疏预填充。

Story heat trend

Current heat 6·Comparable peak 9(Nov 3 18:00)·Comparable change, last 24 hours–

02.557.510Nov3Nov3Nov4Nov4

Trends compare only the same parties observed continuously, so the scope can be narrower than current heat counts. Move the pointer or tap the chart to see hourly heat; use the left and right arrow keys on a keyboard.