<?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>Kernel Sparks</title><link>https://kernelsparks.com/</link><description>Recent content on Kernel Sparks</description><generator>Hugo</generator><language>en</language><lastBuildDate>Wed, 22 Jul 2026 00:00:00 +0300</lastBuildDate><atom:link href="https://kernelsparks.com/index.xml" rel="self" type="application/rss+xml"/><item><title>Why the ASUS GX10 Is the Best GB10 Machine You Can Buy</title><link>https://kernelsparks.com/posts/why-asus-gx10-best-gb10/</link><pubDate>Wed, 22 Jul 2026 00:00:00 +0300</pubDate><guid>https://kernelsparks.com/posts/why-asus-gx10-best-gb10/</guid><description>Every vendor ships the same NVIDIA GB10 Grace Blackwell superchip, so what actually differs? A hands-on comparison of the ASUS Ascent GX10 against the DGX Spark and other GB10 boxes — price, storage, cooling, power delivery, and the details that matter after month three of ownership.</description></item><item><title>About</title><link>https://kernelsparks.com/about/</link><pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate><guid>https://kernelsparks.com/about/</guid><description>&lt;p&gt;I run a small home AI infrastructure built around two ASUS Ascent GX10 (NVIDIA GB10 Grace Blackwell) machines, Macs, routers, and NAS boxes — managed largely through Claude Code and MCP. This blog documents what actually works: measured benchmarks, configs I run in production, and the failure stories in between.&lt;/p&gt;
&lt;p&gt;Contact: &lt;a href="https://x.com/kernelsparks"&gt;@kernelsparks on X&lt;/a&gt;&lt;/p&gt;</description></item></channel></rss>