<?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>Ray on Rudy Pei (裴彦儒)</title>
    <link>https://peabrane.github.io/tags/ray/</link>
    <description>Recent content in Ray on Rudy Pei (裴彦儒)</description>
    <image>
      <title>Rudy Pei (裴彦儒)</title>
      <url>https://peabrane.github.io/profile_pic.jpg</url>
      <link>https://peabrane.github.io/profile_pic.jpg</link>
    </image>
    <generator>Hugo -- 0.163.3</generator>
    <language>en</language>
    <lastBuildDate>Tue, 25 Aug 2026 00:00:00 +0000</lastBuildDate>
    <atom:link href="https://peabrane.github.io/tags/ray/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>KV- and Token-Aware Routing in Ray Serve LLM</title>
      <link>https://peabrane.github.io/post/anyscale-routing/</link>
      <pubDate>Tue, 25 Aug 2026 00:00:00 +0000</pubDate>
      <guid>https://peabrane.github.io/post/anyscale-routing/</guid>
      <description>&lt;p&gt;I coauthored an &lt;a href=&#34;https://www.anyscale.com/blog/llm-kv-token-aware-routing&#34;&gt;Anyscale technical deep dive&lt;/a&gt; on integrating NVIDIA Dynamo&amp;rsquo;s KV-aware selection machinery into Ray Serve LLM. The collaboration is a useful example of a systems component becoming more valuable when it can be embedded outside the system where it started.&lt;/p&gt;</description>
    </item>
  </channel>
</rss>
