<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>PAI-EAS on Chen Kai Blog</title><link>https://www.chenk.top/en/tags/pai-eas/</link><description>Recent content in PAI-EAS on Chen Kai Blog</description><generator>Hugo</generator><language>en</language><lastBuildDate>Sun, 08 Mar 2026 09:00:00 +0000</lastBuildDate><atom:link href="https://www.chenk.top/en/tags/pai-eas/index.xml" rel="self" type="application/rss+xml"/><item><title>Aliyun PAI (4): PAI-EAS — Model Serving, Cold Starts, and the TPS Lie</title><link>https://www.chenk.top/en/aliyun-pai/04-pai-eas-model-serving/</link><pubDate>Sun, 08 Mar 2026 09:00:00 +0000</pubDate><guid>https://www.chenk.top/en/aliyun-pai/04-pai-eas-model-serving/</guid><description>&lt;p>EAS is where the money goes. DSW costs a few hundred RMB a month for development. DLC costs spike. EAS bills 24/7 because someone might call your endpoint, and the &amp;ldquo;minimum replica count&amp;rdquo; in the autoscaler config is the most critical setting in the entire platform. This article covers what I wish I&amp;rsquo;d known before shipping our first production endpoint.&lt;/p>
&lt;p>&lt;figure class="article-figure">
 &lt;img src="https://blog-pic-ck.oss-cn-beijing.aliyuncs.com/posts/en/aliyun-pai/04-pai-eas-model-serving/illustration_1.png" alt="Aliyun PAI (4): PAI-EAS — Model Serving, Cold Starts, and the TPS Lie — Chapter overview" loading="lazy" decoding="async" class="content-image">
 
&lt;/figure>
&lt;/p></description></item></channel></rss>