<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>The Minutes of Dylan Patel on gpu clusters</title>
    <link>https://minutesof.com/dylan-patel/on/gpu-clusters/</link>
    <description>Everything Dylan Patel has said on gpu clusters: 2 verbatim quotes between October 2024 and March 2025, each with a timestamp and a link to the recording…</description>
    <language>en</language>
    <lastBuildDate>Sun, 30 Aug 2026 16:22:23 +0000</lastBuildDate>
    <atom:link href="https://minutesof.com/dylan-patel/on/gpu-clusters/feed.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Patel says current models use 100,000 GPUs while next generation will require hundreds of thousands or million</title>
      <link>https://minutesof.com/q/561bd369-2c88-4ea1-8fa4-3d02389be50d/</link>
      <guid isPermaLink="true">https://minutesof.com/q/561bd369-2c88-4ea1-8fa4-3d02389be50d/</guid>
      <description>“And next generation models that are trained on hundreds of thousands or even millions GPUs, right?” — Special Competitive Studies Project</description>
      <pubDate>Thu, 13 Mar 2025 16:42:15 +0000</pubDate>
      <category>gpu clusters</category>
    </item>
    <item>
      <title>Patel predicts OpenAI will have 300,000 to 500,000 GPU equivalent clusters next year across multiple sites.</title>
      <link>https://minutesof.com/q/5492cf90-3909-4724-b977-ca95780911b3/</link>
      <guid isPermaLink="true">https://minutesof.com/q/5492cf90-3909-4724-b977-ca95780911b3/</guid>
      <description>“Next year, 300 to 500,000 depending on whether it&#x27;s one side or many. Right? 300 to, like, 700,000, I think, is the upper bound of that.” — Dwarkesh Patel</description>
      <pubDate>Wed, 02 Oct 2024 14:59:36 +0000</pubDate>
      <category>gpu clusters</category>
    </item>
  </channel>
</rss>
