<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>gpu on Shubham Srivastava</title><link>https://shubhamsrivastava.co.in/gpu/</link><description>Recent content in gpu on Shubham Srivastava</description><image><title>Shubham Srivastava</title><url>https://shubhamsrivastava.co.in/images/home.png</url><link>https://shubhamsrivastava.co.in/images/home.png</link></image><generator>Hugo -- gohugo.io</generator><language>en-US</language><copyright>Copyright © 2024, Shubham Srivastava.</copyright><lastBuildDate>Tue, 16 Jun 2026 09:00:00 +0530</lastBuildDate><atom:link href="https://shubhamsrivastava.co.in/gpu/index.xml" rel="self" type="application/rss+xml"/><item><title>The Problem With Chasing GPU Utilization</title><link>https://shubhamsrivastava.co.in/blog/the-problem-with-chasing-gpu-utilization/</link><pubDate>Tue, 16 Jun 2026 09:00:00 +0530</pubDate><guid>https://shubhamsrivastava.co.in/blog/the-problem-with-chasing-gpu-utilization/</guid><description>Why maximizing GPU utilization can be the wrong goal, and what metrics actually matter in shared AI infrastructure.</description></item></channel></rss>