<?xml version="1.0" encoding="UTF-8"?>
<rss xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:taxo="http://purl.org/rss/1.0/modules/taxonomy/" version="2.0">
  <channel>
    <title>topic Re: HPUX on Itanium in Operating System - Linux</title>
    <link>https://community.hpe.com/t5/operating-system-linux/hpux-on-itanium/m-p/3874602#M98835</link>
    <description>One thing is that the current DC-Itanium processors are more then twice as fast as the fastest PA-8900 processor. Even is your 32-bit application is not performing 100% on Itanium, the is a certain performance gain.&lt;BR /&gt;If I am not mistaken, you could even try to run your PA-RISC application unmodified on Itanium, using the ARIES emulation software. This way you can probably already get a performance idea without too much effort.&lt;BR /&gt;A native Itanium recompilation will give a much better performance of course.</description>
    <pubDate>Thu, 05 Oct 2006 02:29:34 GMT</pubDate>
    <dc:creator>Wim Rombauts</dc:creator>
    <dc:date>2006-10-05T02:29:34Z</dc:date>
    <item>
      <title>HPUX on Itanium</title>
      <link>https://community.hpe.com/t5/operating-system-linux/hpux-on-itanium/m-p/3874601#M98834</link>
      <description>&lt;!--!*#--&gt;I'm trying to find some info about compiling and running 32-bit applications that use Oracle on an Itanium server running HPUX 11.23.&lt;BR /&gt;&lt;BR /&gt;we currently run on HP PA-RISC servers.  I'm wondering how much of a perfomance change(good or Bad) we will see when we port to the Itanium.  I've been getting conflicting reports on how well Itaniums run 32-bit apps.&lt;BR /&gt;&lt;BR /&gt;We currently aren't ready to port our app to 64-bit but most of our customers prefer HP hardware so we are looking at Itanium.&lt;BR /&gt;&lt;BR /&gt;Thanks for any info or documents that could help me out.&lt;BR /&gt;&lt;BR /&gt;CRS</description>
      <pubDate>Wed, 04 Oct 2006 16:13:37 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-linux/hpux-on-itanium/m-p/3874601#M98834</guid>
      <dc:creator>C R S</dc:creator>
      <dc:date>2006-10-04T16:13:37Z</dc:date>
    </item>
    <item>
      <title>Re: HPUX on Itanium</title>
      <link>https://community.hpe.com/t5/operating-system-linux/hpux-on-itanium/m-p/3874602#M98835</link>
      <description>One thing is that the current DC-Itanium processors are more then twice as fast as the fastest PA-8900 processor. Even is your 32-bit application is not performing 100% on Itanium, the is a certain performance gain.&lt;BR /&gt;If I am not mistaken, you could even try to run your PA-RISC application unmodified on Itanium, using the ARIES emulation software. This way you can probably already get a performance idea without too much effort.&lt;BR /&gt;A native Itanium recompilation will give a much better performance of course.</description>
      <pubDate>Thu, 05 Oct 2006 02:29:34 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-linux/hpux-on-itanium/m-p/3874602#M98835</guid>
      <dc:creator>Wim Rombauts</dc:creator>
      <dc:date>2006-10-05T02:29:34Z</dc:date>
    </item>
    <item>
      <title>Re: HPUX on Itanium</title>
      <link>https://community.hpe.com/t5/operating-system-linux/hpux-on-itanium/m-p/3874603#M98836</link>
      <description>The confusion likely stems from the FUD about running 32-bit _x86_ binaries on Itanium, which is only under Linux and perhaps Windows.&lt;BR /&gt;&lt;BR /&gt;HP-UX provides support for native Itanium 32-bit binaries and they run just fine.  There is no performance penalty for running a native 32-bit Itanium application under HP-UX.&lt;BR /&gt;&lt;BR /&gt;The only wrinkle _might_ be if you link to Oracle libraries - make sure they provide native Itanium 32-bit versions.&lt;BR /&gt;&lt;BR /&gt;Depending on how "clean" your applications are already the "port" to 64-bit might not be all that difficult.  And given that 64-bit has been happening in HP-UX since 1997, it might not be a bad idea to get started :)  There are some options to the HP compilers to get them to emit 64-bit migration warnings.  That would be a good starting point I would think.</description>
      <pubDate>Thu, 05 Oct 2006 11:16:41 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-linux/hpux-on-itanium/m-p/3874603#M98836</guid>
      <dc:creator>rick jones</dc:creator>
      <dc:date>2006-10-05T11:16:41Z</dc:date>
    </item>
    <item>
      <title>Re: HPUX on Itanium</title>
      <link>https://community.hpe.com/t5/operating-system-linux/hpux-on-itanium/m-p/3874604#M98837</link>
      <description>If you don't recompile at all and just copy over the 32bit executables, they will run under the ARIES emulator on Itanium. You can expect to see a truly compute-bound application (no I/O) slow down like in any emulator/interpreter. The general experience is that you'll see about 3x excution rate penalty. But if the application is a database app, the vast majority of clock time is spent waiting on I/O, whether it is getting records from Oracle or waiting for application files. That means that the slower execution time is generally masked by similar I/O time between PARISC and Itanium.&lt;BR /&gt; &lt;BR /&gt;But recompile your 32bit apps on Itanium and yes, you'll see a dramatic increase in performance over PARISC, especially compared to pre-8800 PARISC processors.</description>
      <pubDate>Thu, 05 Oct 2006 11:47:14 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-linux/hpux-on-itanium/m-p/3874604#M98837</guid>
      <dc:creator>Bill Hassell</dc:creator>
      <dc:date>2006-10-05T11:47:14Z</dc:date>
    </item>
    <item>
      <title>Re: HP-UX on Integrity</title>
      <link>https://community.hpe.com/t5/operating-system-linux/hpux-on-itanium/m-p/3874605#M98838</link>
      <description>&lt;P&gt;The difference between +DD32 (the default) and +DD64 is composed of two parts. Using +DD32 causes extra addp4 instructions, possibly increasing instruction cache/TLB. Using +DD64 causes your longs and pointers to double in size, causing data cache/TLB issues.&lt;BR /&gt;&lt;BR /&gt;In most cases, the increase in data size will cause +DD64 to be slower.&lt;BR /&gt;&lt;BR /&gt;&amp;gt;Rick: There are some options to the HP compilers to get them to emit 64-bit migration warnings.&lt;BR /&gt;&lt;BR /&gt;This is +w64bit. Or on PA, +M2.&lt;/P&gt;</description>
      <pubDate>Sat, 29 Oct 2011 08:22:38 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-linux/hpux-on-itanium/m-p/3874605#M98838</guid>
      <dc:creator>Dennis Handly</dc:creator>
      <dc:date>2011-10-29T08:22:38Z</dc:date>
    </item>
  </channel>
</rss>

