<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/">
  <channel>
    <title>本地推理 on AI 热点追踪</title>
    <link>https://ai-news-blog-cg8.pages.dev/tags/%E6%9C%AC%E5%9C%B0%E6%8E%A8%E7%90%86/</link>
    <description>Recent content in 本地推理 on AI 热点追踪</description>
    <generator>Hugo</generator>
    <language>zh-cn</language>
    <lastBuildDate>Sat, 08 Aug 2026 14:40:00 +0800</lastBuildDate>
    <atom:link href="https://ai-news-blog-cg8.pages.dev/tags/%E6%9C%AC%E5%9C%B0%E6%8E%A8%E7%90%86/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>模型越大，部署越小：AI 云的护城河正在裂开吗？</title>
      <link>https://ai-news-blog-cg8.pages.dev/posts/giant-models-tiny-memory-local-inference/</link>
      <pubDate>Sat, 08 Aug 2026 14:40:00 +0800</pubDate>
      <guid>https://ai-news-blog-cg8.pages.dev/posts/giant-models-tiny-memory-local-inference/</guid>
      <description>2.4万亿参数模型与4GB显存推理同时成为热点。大模型正在通过MoE、权重流式加载和分层卸载进入更小设备，但容量、速度与并发是三笔不同的账。</description>
    </item>
  </channel>
</rss>
