<?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: Parallel computing in Intel® Fortran Compiler</title>
    <link>https://community.intel.com/t5/Intel-Fortran-Compiler/Parallel-computing/m-p/854659#M67077</link>
    <description>&lt;DIV style="margin:0px;"&gt;&lt;/DIV&gt;
The goal of the "parallel exploration compiler", which is now part of the product known as Intel Parallel Composer, is to make expressing parallelism easier for programmers who are not used to the concept. At present, the parallel directives added are just macros for OpenMP directives, though that may change in the future. This concept doesn't apply well to Fortran syntax which does not have lexical nested blocks. Anything you can do with the parallel extensions in C++ you can do in Fortran with OpenMP.&lt;BR /&gt;&lt;BR /&gt;Depending on how successful the various features are, we may look to see how they could be applied to Fortran, but in my experience, Fortran programmers are more comfortable with OpenMP and the benefit seems less.&lt;BR /&gt;</description>
    <pubDate>Thu, 29 Jan 2009 20:14:47 GMT</pubDate>
    <dc:creator>Steven_L_Intel1</dc:creator>
    <dc:date>2009-01-29T20:14:47Z</dc:date>
    <item>
      <title>Parallel computing</title>
      <link>https://community.intel.com/t5/Intel-Fortran-Compiler/Parallel-computing/m-p/854658#M67076</link>
      <description>&lt;BR /&gt;Hi&lt;BR /&gt;&lt;BR /&gt;I look like Intel is promoting the product Intel  C + + Parallel Exploration Compiler, Prototype Edition&lt;BR /&gt;&lt;BR /&gt;Is there any development in this direction for fortran?&lt;BR /&gt;&lt;BR /&gt;Does anyone have any news?&lt;BR /&gt;</description>
      <pubDate>Thu, 29 Jan 2009 19:47:20 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-Fortran-Compiler/Parallel-computing/m-p/854658#M67076</guid>
      <dc:creator>Luis2</dc:creator>
      <dc:date>2009-01-29T19:47:20Z</dc:date>
    </item>
    <item>
      <title>Re: Parallel computing</title>
      <link>https://community.intel.com/t5/Intel-Fortran-Compiler/Parallel-computing/m-p/854659#M67077</link>
      <description>&lt;DIV style="margin:0px;"&gt;&lt;/DIV&gt;
The goal of the "parallel exploration compiler", which is now part of the product known as Intel Parallel Composer, is to make expressing parallelism easier for programmers who are not used to the concept. At present, the parallel directives added are just macros for OpenMP directives, though that may change in the future. This concept doesn't apply well to Fortran syntax which does not have lexical nested blocks. Anything you can do with the parallel extensions in C++ you can do in Fortran with OpenMP.&lt;BR /&gt;&lt;BR /&gt;Depending on how successful the various features are, we may look to see how they could be applied to Fortran, but in my experience, Fortran programmers are more comfortable with OpenMP and the benefit seems less.&lt;BR /&gt;</description>
      <pubDate>Thu, 29 Jan 2009 20:14:47 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-Fortran-Compiler/Parallel-computing/m-p/854659#M67077</guid>
      <dc:creator>Steven_L_Intel1</dc:creator>
      <dc:date>2009-01-29T20:14:47Z</dc:date>
    </item>
    <item>
      <title>Re: Parallel computing</title>
      <link>https://community.intel.com/t5/Intel-Fortran-Compiler/Parallel-computing/m-p/854660#M67078</link>
      <description>&lt;DIV style="margin:0px;"&gt;&lt;/DIV&gt;
&lt;BR /&gt;Steve,&lt;BR /&gt;&lt;BR /&gt;&amp;gt;&amp;gt; This concept doesn't apply well to Fortran syntax which does not have lexical nested blocks.&lt;BR /&gt;&lt;BR /&gt;I know it is not quite the same as C/C++ {}'s but by using !$OMP PARALLEL... !$OMP END PARALLEL you "sort of" get the equivilent of {}'s as you can declare what variables have local scope (they have to be declared outside of scope prior to use inside of scope). &lt;BR /&gt;&lt;BR /&gt;Jim&lt;BR /&gt;</description>
      <pubDate>Thu, 29 Jan 2009 23:44:12 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-Fortran-Compiler/Parallel-computing/m-p/854660#M67078</guid>
      <dc:creator>jimdempseyatthecove</dc:creator>
      <dc:date>2009-01-29T23:44:12Z</dc:date>
    </item>
    <item>
      <title>Re: Parallel computing</title>
      <link>https://community.intel.com/t5/Intel-Fortran-Compiler/Parallel-computing/m-p/854661#M67079</link>
      <description>&lt;DIV style="margin:0px;"&gt;&lt;/DIV&gt;
Sure, but you're then creating a parallel region (which doesn't have to be lexical in nature - calls to other routines from within a parallel region are parallel.)&lt;BR /&gt;&lt;BR /&gt;The focus of Parallel Composer is to entice so-called "mainstream programmers" to add parallelism to their applications. It does not add any capability that won't also be in Intel Fortran, except one I suppose - a parallel debugger extension which relies on some additional debug information from the compiler. There's no reason it couldn't be added to Fortran as well.&lt;BR /&gt;</description>
      <pubDate>Fri, 30 Jan 2009 01:57:51 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-Fortran-Compiler/Parallel-computing/m-p/854661#M67079</guid>
      <dc:creator>Steven_L_Intel1</dc:creator>
      <dc:date>2009-01-30T01:57:51Z</dc:date>
    </item>
  </channel>
</rss>

