<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>RDMA on Rik Kisnah - Blog</title><link>https://www.rik-kisnah.ai/tags/rdma/</link><description>Recent content in RDMA on Rik Kisnah - Blog</description><generator>Hugo</generator><language>en</language><lastBuildDate>Sun, 26 Oct 2025 22:43:15 -0400</lastBuildDate><atom:link href="https://www.rik-kisnah.ai/tags/rdma/feed.xml" rel="self" type="application/rss+xml"/><item><title>From First Principles to Zettascale: How OCI's GPU/RDMA Architecture Redefines AI Infrastructure</title><link>https://www.rik-kisnah.ai/posts/summary-gpu-oci-first-principles-blog/</link><pubDate>Sun, 26 Oct 2025 22:43:15 -0400</pubDate><guid>https://www.rik-kisnah.ai/posts/summary-gpu-oci-first-principles-blog/</guid><description>Disclaimer: This article reflects my personal research and analysis based on publicly available information and is not representative of my employer&amp;rsquo;s official position.
In the rapidly evolving landscape of AI infrastructure, one company has quietly revolutionized how we think about GPU computing at scale. Through a series of &amp;ldquo;First Principles&amp;rdquo; engineering blogs and groundbreaking deployments, Oracle Cloud Infrastructure (OCI) has demonstrated that starting from fundamental physics and systems design—rather than following industry conventions—can yield extraordinary results.</description></item></channel></rss>