<?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>Episode-17 on GPU 5090 — devlog</title>
    <link>https://gpu.local-xyz.ru/blog/tags/episode-17/</link>
    <description>Recent content in Episode-17 on GPU 5090 — devlog</description>
    <generator>Hugo</generator>
    <language>ru-ru</language>
    <lastBuildDate>Thu, 07 May 2026 12:12:00 +0000</lastBuildDate>
    <atom:link href="https://gpu.local-xyz.ru/blog/tags/episode-17/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Episode #17 — production cycle reality, narration format ROI</title>
      <link>https://gpu.local-xyz.ru/blog/posts/2026-05-07-day-13-alpha-episode-17/</link>
      <pubDate>Thu, 07 May 2026 12:12:00 +0000</pubDate>
      <guid>https://gpu.local-xyz.ru/blog/posts/2026-05-07-day-13-alpha-episode-17/</guid>
      <description>Третий narration episode подряд. Technical reflection: pure 4DGS narration pipeline = ~15 sec compute per episode (vs ~5 min talking-head v7 LS pipeline = 20× faster). Trade-off — body статика, но 4D rotation сохраняется. Acceptable для voice-over content. Sustained cadence proven viable.</description>
    </item>
  </channel>
</rss>
