
  <rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
    <channel>
      <title>Shivanshu Pathak · Software Engineer, Trading Systems</title>
      <link>https://shivanshu814.vercel.app/blog</link>
      <description>Shivanshu Pathak: software engineer building trading systems. Pre-trade risk and order execution for a derivatives venue, audit-grade financial data pipelines, and low-latency C++ work in progress.</description>
      <language>en-us</language>
      <managingEditor>shivanshu814pathak@gmail.com (Shivanshu Pathak)</managingEditor>
      <webMaster>shivanshu814pathak@gmail.com (Shivanshu Pathak)</webMaster>
      <lastBuildDate>Tue, 17 Nov 2026 00:00:00 GMT</lastBuildDate>
      <atom:link href="https://shivanshu814.vercel.app/tags/low-latency/feed.xml" rel="self" type="application/rss+xml"/>
      
  <item>
    <guid>https://shivanshu814.vercel.app/blog/engineering/order-book-not-std-map</guid>
    <title>Your order book should not be a std::map</title>
    <link>https://shivanshu814.vercel.app/blog/engineering/order-book-not-std-map</link>
    <description>A flat, price-indexed order book with intrusive FIFO queues and pooled orders against the textbook std::map design: same workload, same fills, 1.9× the throughput. Includes the numbers, the caveats, and where the time actually goes.</description>
    <pubDate>Tue, 17 Nov 2026 00:00:00 GMT</pubDate>
    <author>shivanshu814pathak@gmail.com (Shivanshu Pathak)</author>
    <category>cpp</category><category>low-latency</category><category>order-book</category><category>performance</category>
  </item>

  <item>
    <guid>https://shivanshu814.vercel.app/blog/engineering/spsc-queue-latency-vs-throughput</guid>
    <title>My lock-free queue lost on throughput. It still won.</title>
    <link>https://shivanshu814.vercel.app/blog/engineering/spsc-queue-latency-vs-throughput</link>
    <description>An SPSC ring buffer with acquire/release ordering and cache-line padding: 1.7× slower than a mutex queue on raw throughput, but a 3.7× faster median round trip and a 20× better p99. Why both results are correct, and what that means for trading systems.</description>
    <pubDate>Mon, 16 Nov 2026 00:00:00 GMT</pubDate>
    <author>shivanshu814pathak@gmail.com (Shivanshu Pathak)</author>
    <category>cpp</category><category>concurrency</category><category>lock-free</category><category>low-latency</category>
  </item>

    </channel>
  </rss>
