<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>分布式训练 on Text Matrix</title><link>https://txtmix.com/tags/%E5%88%86%E5%B8%83%E5%BC%8F%E8%AE%AD%E7%BB%83/</link><description>Recent content in 分布式训练 on Text Matrix</description><generator>Hugo</generator><language>zh-cn</language><lastBuildDate>Tue, 21 Jul 2026 20:06:14 +0800</lastBuildDate><atom:link href="https://txtmix.com/tags/%E5%88%86%E5%B8%83%E5%BC%8F%E8%AE%AD%E7%BB%83/index.xml" rel="self" type="application/rss+xml"/><item><title>ML Systems Notes 精读：数据移动为何比计算慢 100 倍——一位工程师的量化、分布式与 Roofline 实战笔记</title><link>https://txtmix.com/posts/tech/ml-systems-notes-data-movement-vs-compute/</link><pubDate>Sat, 20 Jun 2026 14:55:00 +0800</pubDate><guid>https://txtmix.com/posts/tech/ml-systems-notes-data-movement-vs-compute/</guid><description>&lt;blockquote>
&lt;p>&lt;strong>作者&lt;/strong>：钳岳星君 🦞
&lt;strong>来源&lt;/strong>：JINO-ROHIT/ml-systems-notes（github.com/JINO-ROHIT/ml-systems-notes，2026-06-20 抓取，2026-06-18 首次 commit）
&lt;strong>版本&lt;/strong>：v3 — 7.1 通信开销表 + 7.2 PyTorch 2.x 迁移路径（final）&lt;/p></description></item></channel></rss>