<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Retro-Tech on SmartStack: AI, Self-Hosting &amp; Smart Finance</title><link>https://www.smart-stacking.com/tags/retro-tech/</link><description>Recent content in Retro-Tech on SmartStack: AI, Self-Hosting &amp; Smart Finance</description><generator>Hugo</generator><language>en</language><copyright>&amp;copy; 2026 SmartStack</copyright><lastBuildDate>Sat, 05 Sep 2026 12:00:03 +0800</lastBuildDate><atom:link href="https://www.smart-stacking.com/tags/retro-tech/index.xml" rel="self" type="application/rss+xml"/><item><title>Running a 90M LLM on a Sony PSP: Nostalgia and Overkill in 2026</title><link>https://www.smart-stacking.com/posts/2026-09-05-you-can-now-run-a-90m-conversational-llm-on-the-sony-psp-hardware-from-2004-doesnt-get-more-local-than-this/</link><pubDate>Sat, 05 Sep 2026 12:00:03 +0800</pubDate><guid>https://www.smart-stacking.com/posts/2026-09-05-you-can-now-run-a-90m-conversational-llm-on-the-sony-psp-hardware-from-2004-doesnt-get-more-local-than-this/</guid><description>Someone ran a 90M LLM on a Sony PSP (yes, the 2004 one). It&amp;rsquo;s ridiculous, brilliant, and says a lot about where AI hardware is heading.</description></item></channel></rss>