<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Revec: Program Rejuvenation through Revectorization · HackerPapers discussions</title><link>https://hackerpaper.kaipfister.de/p/1902.02816/</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/1902.02816/feed.xml" rel="self" type="application/rss+xml" /><item><title>Revec: Program Rejuvenation through Revectorization — mentioned by ajayjain</title><link>https://news.ycombinator.com/item?id=23824191</link><guid isPermaLink="false">urn:hackerpapers:mention:1902.02816:23824191</guid><pubDate>Mon, 13 Jul 2020 18:50:02 GMT</pubDate><description>&lt;p&gt;Two years ago, I wrote an LLVM compiler pass that automatically upgrades SSE or AVX-2 code to AVX-512 when those instructions are available. It&amp;#x27;s helpful for getting performance gains for hand engineered code that you don&amp;#x27;t want to touch. We saw some good gains on integer workloads (unpacking, useful for databases - something I&amp;#x27;d call &amp;quot;regular code&amp;quot;): 1.16x speedup going from SSE to AVX-2, and 1.43x speedup going from SSE to AVX-512. Even better speedups are available for FP workloads, though hand-writing the AVX-512 version can work a lot better since programmers can exploit the full range…&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/1902.02816/"&gt;Paper and discussion history&lt;/a&gt;. A mention is not an endorsement.&lt;/p&gt;</description><category>cs.PL</category></item><item><title>Revec: Program Rejuvenation through Revectorization — HN submission</title><link>https://news.ycombinator.com/item?id=20555874</link><guid isPermaLink="false">urn:hackerpapers:submission:1902.02816:20555874</guid><pubDate>Mon, 29 Jul 2019 15:53:00 GMT</pubDate><description>&lt;p&gt;Revec: Program Rejuvenation Through Revectorization&lt;/p&gt;&lt;p&gt;&lt;/p&gt;&lt;p&gt;1 point; 0 discussion replies on HN. &lt;a href="https://hackerpaper.kaipfister.de/p/1902.02816/"&gt;Paper and source history&lt;/a&gt;. A submission is not an endorsement.&lt;/p&gt;</description><category>cs.PL</category></item><item><title>Revec: Program Rejuvenation through Revectorization — HN submission</title><link>https://news.ycombinator.com/item?id=20545648</link><guid isPermaLink="false">urn:hackerpapers:submission:1902.02816:20545648</guid><pubDate>Sun, 28 Jul 2019 03:12:43 GMT</pubDate><description>&lt;p&gt;Revec: Program Rejuvenation Through Revectorization&lt;/p&gt;&lt;p&gt;&lt;/p&gt;&lt;p&gt;2 points; 0 discussion replies on HN. &lt;a href="https://hackerpaper.kaipfister.de/p/1902.02816/"&gt;Paper and source history&lt;/a&gt;. A submission is not an endorsement.&lt;/p&gt;</description><category>cs.PL</category></item></channel></rss>
