English(EN) One Laptop, Eight Models: The Dense Rows Read Memory at 89% of the Ceiling, the MoE Rows at 43-67%

本地LLM基准测试显示MoE模型在内存带宽方面存在困难

作者 PulseAugur 编辑部 · [1 个来源] · 2026-11-15 03:02

一位用户在Surface Laptop Ultra上使用llama.cpp在本地运行了八个不同的大型语言模型（LLM）并进行了基准测试。测试重点关注内存带宽，结果显示，稠密模型达到了系统理论内存天花板的约89%，而混合专家（MoE）模型仅达到43%-67%。这一显著的性能差距凸显了MoE架构中路由机制带来的开销。 AI

影响凸显了MoE架构在内存受限场景下的性能成本。

排序理由用户生成的关于LLM在特定硬件上性能的基准测试分析。[lever_c_demoted from research: ic=1 ai=0.7]

在 dev.to — LLM tag 阅读 →

基础设施

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

本文如何被排名

Signal score: 28 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 用户生成的关于LLM在特定硬件上性能的基准测试分析。[lever_c_demoted from research: ic=1 ai=0.7]
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; 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]

dev.to — LLM tag TIER_1 English(EN) · Jax Sinclair · 2026-11-15 03:02

一台笔记本，八个模型：密集行读取内存占用率达天花板的89%，MoE行占用率在43-67%

<p>A reader with a Surface Laptop Ultra sent me a table of eight local LLMs on the same machine, same llama.cpp build, same commands, speculative decoding off. His numbers are generation speed, which is a memory-bandwidth result if you treat it right.</p> <p>So I treated it right…

本文如何被排名

报道来源 [1]

一台笔记本，八个模型：密集行读取内存占用率达天花板的89%，MoE行占用率在43-67%

相关实体

相关话题