PulseAugur
实时 06:28:38
English(EN) DynaNDE: Dynamic Near-Data Expert Scheduling for Batched MoE Inference

DynaNDE框架通过NPU-NDP调度提升MoE推理速度

研究人员开发了DynaNDE,一个动态调度框架,旨在加速基于神经网络处理单元(NPU)的系统上的批处理专家混合(MoE)模型推理。该框架利用近数据处理(NDP)来减少数据移动开销,这是MoE模型的一个主要瓶颈。DynaNDE包含一个分析性能模型,用于优化跨NPU和NDP的专家调度,并考虑硬件异构性和并发性。实验结果表明,与现有的最先进方法相比,DynaNDE在预填充阶段可实现2.6倍的平均加速,在解码阶段可实现2.2倍的平均加速。 AI

影响 优化MoE推理效率,可能降低大型语言模型的硬件成本和延迟。

排序理由 这是一篇详细介绍用于优化AI模型推理的新框架的研究论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.LG 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。 我们如何撰写摘要 →

DynaNDE框架通过NPU-NDP调度提升MoE推理速度

本文如何被排名

Signal score
30 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
这是一篇详细介绍用于优化AI模型推理的新框架的研究论文。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration
Single-source cluster
Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics
infra, paper
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
Breaking (< 6h)
Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

完整方法见我们的编辑标准

报道来源 [1]

  1. arXiv cs.LG TIER_1 English(EN) · Xiaoyang Lu, Belthangady Akash Vi Narayana Pai, Xian-He Sun ·

    DynaNDE:动态近数据专家调度用于批处理MoE推理

    arXiv:2609.00407v1 Announce Type: cross Abstract: Mixture-of-Experts (MoE) models enable efficient scaling of large language model (LLM) inference but suffer from substantial data-movement overhead when deployed on neural processing unit (NPU)-based systems. Near-Data Processing …