<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>TResNet: High Performance GPU-Dedicated Architecture · HackerPapers discussions</title><link>https://hackerpaper.kaipfister.de/p/2003.13630/</link><description>Collected HN comments and submissions linking this paper.</description><language>en</language><lastBuildDate>Mon, 02 Nov 2026 11:16:22 GMT</lastBuildDate><atom:link href="https://hackerpaper.kaipfister.de/p/2003.13630/feed.xml" rel="self" type="application/rss+xml" /><item><title>TResNet: High Performance GPU-Dedicated Architecture — mentioned by CompleteSkeptic</title><link>https://news.ycombinator.com/item?id=24045563</link><guid isPermaLink="false">urn:hackerpapers:mention:2003.13630:24045563</guid><pubDate>Tue, 04 Aug 2020 01:29:28 GMT</pubDate><description>&lt;p&gt;There actually is some work (https://arxiv.org/abs/2003.13630) claiming that FLOPS are a poor measure of real-world performance - with some of the more recent FLOP-efficient models actually running slower than older models.&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/2003.13630/"&gt;Paper and discussion history&lt;/a&gt;. A mention is not an endorsement.&lt;/p&gt;</description><category>cs.CV</category></item></channel></rss>
