<?xml version="1.0" encoding="utf-8"?>
<feed xml:lang="en" xmlns="http://www.w3.org/2005/Atom"><title>Recent changes to 50: Improve performance of the emulation on PowerPC 970 (G5)</title><link href="https://sourceforge.net/p/euaeppcjit/tickets/50/" rel="alternate"/><link href="https://sourceforge.net/p/euaeppcjit/tickets/50/feed.atom" rel="self"/><id>https://sourceforge.net/p/euaeppcjit/tickets/50/</id><updated>2014-06-21T23:26:29.486000Z</updated><subtitle>Recent changes to 50: Improve performance of the emulation on PowerPC 970 (G5)</subtitle><entry><title>Improve performance of the emulation on PowerPC 970 (G5)</title><link href="https://sourceforge.net/p/euaeppcjit/tickets/50/" rel="alternate"/><published>2014-06-21T23:26:29.486000Z</published><updated>2014-06-21T23:26:29.486000Z</updated><author><name>Almos Rajnai</name><uri>https://sourceforge.net/u/racsdragon/</uri></author><id>https://sourceforge.netd3b1aa162191316fb9be04dcbe79df60455360ce</id><summary type="html">&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;</summary></entry></feed>