<?xml version="1.0" encoding="utf-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Recent changes to 50: Improve performance of the emulation on PowerPC 970 (G5)</title><link>https://sourceforge.net/p/euaeppcjit/tickets/50/</link><description>Recent changes to 50: Improve performance of the emulation on PowerPC 970 (G5)</description><atom:link href="https://sourceforge.net/p/euaeppcjit/tickets/50/feed.rss" rel="self"/><language>en</language><lastBuildDate>Sat, 21 Jun 2014 23:26:29 -0000</lastBuildDate><atom:link href="https://sourceforge.net/p/euaeppcjit/tickets/50/feed.rss" rel="self" type="application/rss+xml"/><item><title>Improve performance of the emulation on PowerPC 970 (G5)</title><link>https://sourceforge.net/p/euaeppcjit/tickets/50/</link><description>&lt;div class="markdown_content"&gt;&lt;p&gt;As it is perceived currently: the compiled code does not perform better on PowerPC 970 (aka G5) than on the older processors.&lt;br /&gt;
PPC 970 requires special optimization strategies which are not resolved in the current JIT compiler.&lt;/p&gt;
&lt;p&gt;Possible areas for improvement:&lt;/p&gt;
&lt;ul&gt;
&lt;li&gt;
&lt;p&gt;set up the branch hinting properly, so the branch can be predicted more accurately&lt;/p&gt;
&lt;/li&gt;
&lt;li&gt;
&lt;p&gt;make use of multiple cores&lt;/p&gt;
&lt;/li&gt;
&lt;li&gt;
&lt;p&gt;revise the emitted code in respect for the multiple execution units and make use of out-of-order parallel execution as much as possible&lt;/p&gt;
&lt;/li&gt;
&lt;li&gt;
&lt;p&gt;revise the usage of various PPC instructions, if there was a better choice for PPC970 which is faster&lt;/p&gt;
&lt;/li&gt;
&lt;/ul&gt;&lt;/div&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Almos Rajnai</dc:creator><pubDate>Sat, 21 Jun 2014 23:26:29 -0000</pubDate><guid>https://sourceforge.netd3b1aa162191316fb9be04dcbe79df60455360ce</guid></item></channel></rss>