<?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>Rdma on Route179</title>
    <link>https://route179.dev/tags/rdma/</link>
    <description>Recent content in Rdma on Route179</description>
    <generator>Hugo</generator>
    <language>en-us</language>
    <copyright>2026 Sheng Chen</copyright>
    <lastBuildDate>Tue, 21 Jul 2026 10:00:00 +1000</lastBuildDate>
    <atom:link href="https://route179.dev/tags/rdma/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>DGX Spark Clustering - RoCE/RDMA Networking Setup &amp; NCCL Benchmarking</title>
      <link>https://route179.dev/2026/07/21/dgx-spark-nccl-roce-benchmarking/</link>
      <pubDate>Tue, 21 Jul 2026 10:00:00 +1000</pubDate>
      <guid>https://route179.dev/2026/07/21/dgx-spark-nccl-roce-benchmarking/</guid>
      <description>&lt;p&gt;I&amp;rsquo;ve been running a single DGX Spark (GB10) as an EKS Hybrid Node for a while, and
recently a second one landed on my desk. The plan is to stack the two Sparks with one QSFP112 cable
to create a little two-node EKS Hybrid cluster for distributed inference.
This post documents how to set up RoCE/RDMA networking and perform NCCL testing across the two Spark nodes, including all lessons learned.&lt;/p&gt;</description>
    </item>
  </channel>
</rss>
