<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>AGI基础设施 on AI内参</title>
    <link>https://www.neican.ai/tags/agi%E5%9F%BA%E7%A1%80%E8%AE%BE%E6%96%BD/</link>
    <description>Recent content in AGI基础设施 on AI内参</description>
    <generator>Hugo</generator>
    <language>zh-cn</language>
    <lastBuildDate>Mon, 05 Jan 2026 16:40:02 +0800</lastBuildDate>
    <atom:link href="https://www.neican.ai/tags/agi%E5%9F%BA%E7%A1%80%E8%AE%BE%E6%96%BD/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>算力公式的“减法”逻辑：专访无问芯穹夏立雪，解码AI下半场的生死认知</title>
      <link>https://www.neican.ai/insights/article-20260105164002793-2/</link>
      <pubDate>Mon, 05 Jan 2026 16:40:02 +0800</pubDate>
      <guid>https://www.neican.ai/insights/article-20260105164002793-2/</guid>
      <description>本文深度专访了无问芯穹联合创始人兼CEO夏立雪，探讨了其如何通过软硬协同优化技术解决AI算力瓶颈。夏立雪指出，2025年AI行业将从“造模型”转向“高效用模型”阶段，企业间的差距将由“一号位”的AI认知度决定，同时中国开源模型正主导全球应用。</description>
    </item>
    <item>
      <title>“AI工厂”：通往AGI的系统级架构，而非算力暴力美学</title>
      <link>https://www.neican.ai/insights/aiagi-20250730111005253-1/</link>
      <pubDate>Wed, 30 Jul 2025 11:10:05 +0800</pubDate>
      <guid>https://www.neican.ai/insights/aiagi-20250730111005253-1/</guid>
      <description>摩尔线程提出的“AI工厂”理念，突破了传统算力堆叠的局限，通过集成全功能GPU、深度优化软硬件协同及构建高效率高稳定性的集群系统，为实现通用人工智能（AGI）提供了系统级工程方案。这一创新不仅显著提升了AI训练与推理的效率，加速了国产化进程，更在商业落地与前沿科学探索间架起桥梁，预示着AI基础设施从“暴力美学”向“智慧生产”的范式转变。</description>
    </item>
  </channel>
</rss>
