<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Compilers on Ryan Orban</title><link>https://ryanorban.com/categories/compilers/</link><description>Recent content in Compilers on Ryan Orban</description><generator>Hugo</generator><language>en-us</language><managingEditor>me@ryanorban.com (Ryan Orban)</managingEditor><webMaster>me@ryanorban.com (Ryan Orban)</webMaster><copyright>Ryan Orban</copyright><lastBuildDate>Thu, 24 Feb 2022 00:00:00 +0000</lastBuildDate><atom:link href="https://ryanorban.com/categories/compilers/index.xml" rel="self" type="application/rss+xml"/><item><title>Dive into Deep Learning Compiler</title><link>https://ryanorban.com/notes/dive-into-deep-learning-compiler/</link><pubDate>Thu, 24 Feb 2022 00:00:00 +0000</pubDate><author>me@ryanorban.com (Ryan Orban)</author><guid>https://ryanorban.com/notes/dive-into-deep-learning-compiler/</guid><description>&lt;h3 id="summary" class="scroll-mt-8 group"&gt;
 Summary
 
 &lt;a href="#summary"
 class="no-underline hidden opacity-50 hover:opacity-100 !text-inherit group-hover:inline-block"
 aria-hidden="true" title="Link to this heading" tabindex="-1"&gt;
 &lt;svg
 xmlns="http://www.w3.org/2000/svg"
 width="16"
 height="16"
 fill="none"
 stroke="currentColor"
 stroke-linecap="round"
 stroke-linejoin="round"
 stroke-width="2"
 class="lucide lucide-link w-4 h-4 block"
 viewBox="0 0 24 24"
&gt;
 &lt;path d="M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72 1.71" /&gt;
 &lt;path d="M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71" /&gt;
&lt;/svg&gt;

 &lt;/a&gt;
 
&lt;/h3&gt;
&lt;p&gt;&lt;em&gt;Dive into Deep Learning Compiler&lt;/em&gt; is the companion textbook for Apache TVM, the open-source deep learning compiler framework. It covers the full compilation pipeline: how a high-level model defined in PyTorch or TensorFlow gets lowered through multiple intermediate representations (IRs), optimized, and finally compiled to machine code for CPUs, GPUs, or custom AI accelerators.&lt;/p&gt;</description></item><item><title>Sol: A Sunny Little Virtual Machine</title><link>https://ryanorban.com/notes/sol-virtual-machine/</link><pubDate>Mon, 15 Oct 2012 00:00:00 +0000</pubDate><author>me@ryanorban.com (Ryan Orban)</author><guid>https://ryanorban.com/notes/sol-virtual-machine/</guid><description>&lt;p&gt;&lt;img
 src="https://ryanorban.com/images/notes/sol-virtual-machine.png"
 alt="Sol: A Sunny Little Virtual Machine" class="note-hero-img"
 loading="lazy"
/&gt;
&lt;/p&gt;
&lt;h3 id="summary" class="scroll-mt-8 group"&gt;
 Summary
 
 &lt;a href="#summary"
 class="no-underline hidden opacity-50 hover:opacity-100 !text-inherit group-hover:inline-block"
 aria-hidden="true" title="Link to this heading" tabindex="-1"&gt;
 &lt;svg
 xmlns="http://www.w3.org/2000/svg"
 width="16"
 height="16"
 fill="none"
 stroke="currentColor"
 stroke-linecap="round"
 stroke-linejoin="round"
 stroke-width="2"
 class="lucide lucide-link w-4 h-4 block"
 viewBox="0 0 24 24"
&gt;
 &lt;path d="M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72 1.71" /&gt;
 &lt;path d="M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71" /&gt;
&lt;/svg&gt;

 &lt;/a&gt;
 
&lt;/h3&gt;
&lt;p&gt;Rasmus Andersson (creator of the Figma typeface system and various open-source tools) built Sol as an educational process virtual machine designed to run inside an OS process. The name means sun in Swedish. Where most toy VMs are stack-based for simplicity, Sol uses a register machine design — the same approach taken by Lua&amp;rsquo;s VM and CPython&amp;rsquo;s modern bytecode interpreter — on the grounds that registers reduce opcode count and improve execution speed despite slightly larger instruction encoding.&lt;/p&gt;</description></item></channel></rss>