<?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>Mixed Precision on a.dev</title><link>https://aditya08.github.io/tags/mixed-precision/</link><description>Recent content in Mixed Precision on a.dev</description><generator>Hugo -- 0.148.0</generator><language>en</language><lastBuildDate>Tue, 16 Jun 2026 00:00:00 +0000</lastBuildDate><atom:link href="https://aditya08.github.io/tags/mixed-precision/index.xml" rel="self" type="application/rss+xml"/><item><title>Mixed-Precision Communication-Avoiding SGD for Generalized Linear Models on GPUs</title><link>https://aditya08.github.io/publications/mixedprec/</link><pubDate>Tue, 16 Jun 2026 00:00:00 +0000</pubDate><guid>https://aditya08.github.io/publications/mixedprec/</guid><description>This paper presents a mixed-precision communication-avoiding SGD method for generalized linear models that exploits reduced-precision matrix hardware on GPUs</description></item></channel></rss>