<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Local-Llm on Danilo Falcão da Silva</title><link>https://falcao.org/tags/local-llm/</link><description>Recent content in Local-Llm on Danilo Falcão da Silva</description><generator>Hugo</generator><language>en-us</language><lastBuildDate>Wed, 03 Jun 2026 20:04:52 -0300</lastBuildDate><atom:link href="https://falcao.org/tags/local-llm/index.xml" rel="self" type="application/rss+xml"/><item><title>NVIDIA Spark vs AMD Ryzen AI Max: Which Little AI Box Do I Actually Want?</title><link>https://falcao.org/posts/nvidia-spark-vs-amd-ai-max/</link><pubDate>Wed, 03 Jun 2026 20:04:52 -0300</pubDate><guid>https://falcao.org/posts/nvidia-spark-vs-amd-ai-max/</guid><description>&lt;p>For two years the answer to &amp;ldquo;where do I run a model that won&amp;rsquo;t fit on a
normal GPU&amp;rdquo; was the same boring answer: rent a cloud instance, watch the
meter, hope you remembered to shut it down. In 2026 there&amp;rsquo;s finally a
second answer that lives on your desk. Two of them, actually, and they
take opposite philosophical bets.&lt;/p>
&lt;p>NVIDIA shrank a slice of its datacenter stack into a desktop and called
it &lt;strong>DGX Spark&lt;/strong> — a GB10 Grace Blackwell superchip with 128GB of unified
memory and a literal petaFLOP of FP4 compute. AMD took the other road:
&lt;strong>Ryzen AI Max+ 395&lt;/strong>, codename Strix Halo, a normal x86 APU that happens
to carry up to 128GB of unified memory and shows up in mini PCs you can
buy for roughly half the price.&lt;/p></description></item></channel></rss>