<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Pde on Dmitry Serikov</title><link>https://dmitryserikoff.com/tags/pde/</link><description>Recent content in Pde on Dmitry Serikov</description><generator>Hugo</generator><language>en-us</language><lastBuildDate>Thu, 09 Jul 2026 00:00:00 +0300</lastBuildDate><atom:link href="https://dmitryserikoff.com/tags/pde/index.xml" rel="self" type="application/rss+xml"/><item><title>The Price of an Answer</title><link>https://dmitryserikoff.com/posts/the-price-of-an-answer/</link><pubDate>Thu, 09 Jul 2026 00:00:00 +0300</pubDate><guid>https://dmitryserikoff.com/posts/the-price-of-an-answer/</guid><description>&lt;p>&amp;ldquo;Nonlinear means no shortcuts&amp;rdquo; is one of those phrases everyone repeats and
nobody measures. So I measured it. Take 35 famous nonlinear PDEs — Burgers,
KdV, chaos, solitons — and ask: as parameters vary, how many independent
directions do the solutions actually explore?&lt;/p>
&lt;p>The diagnostic is &lt;strong>effective rank&lt;/strong> (Φ₁). The result: 33 of the 35 equations
sit at Φ₁ ≈ 1. Their parametric solution families trace a simple
one-dimensional curve through high-dimensional space — walking along a wire.
Which means about seven strategic solves are enough for machine-precision
forward prediction, inverse parameter recovery, and field reconstruction, all
in microseconds.&lt;/p></description></item></channel></rss>