<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"
	xmlns:content="http://purl.org/rss/1.0/modules/content/"
	xmlns:dc="http://purl.org/dc/elements/1.1/"
	xmlns:atom="http://www.w3.org/2005/Atom"
	>
<channel>
	<title>Comments on: Choice Overload and Parallel Programming</title>
	<atom:link href="http://www.thinkingparallel.com/2007/10/18/choice-overload-and-parallel-programming/feed/" rel="self" type="application/rss+xml" />
	<link>http://www.thinkingparallel.com/2007/10/18/choice-overload-and-parallel-programming/</link>
	<description>A Blog on Parallel Programming and Concurrency by Michael Suess</description>
	<pubDate>Wed, 09 Jul 2008 02:34:47 +0000</pubDate>
	<generator>http://wordpress.org/?v=2.5</generator>
		<item>
		<title>By: Jakob Engblom</title>
		<link>http://www.thinkingparallel.com/2007/10/18/choice-overload-and-parallel-programming/#comment-32667</link>
		<dc:creator>Jakob Engblom</dc:creator>
		<pubDate>Fri, 23 Nov 2007 11:27:24 +0000</pubDate>
		<guid isPermaLink="false">http://www.thinkingparallel.com/2007/10/18/choice-overload-and-parallel-programming/#comment-32667</guid>
		<description>I agree that we need new languages, but I have a feeling that the successful ones will be those that are higher-level and more tailored to a domain. I talked to Jim on this in August this year, and since his background is mostly in HPC, he does not really have that kind of experience.  He thinks in "coding" languages like C/Fortran/Erlang. 

I commented on this here: http://jakob.engbloms.se/archives/35#more-35 and here: http://jakob.engbloms.se/archives/55#more-55 . Anyone interested should check out what Labview is doing in parallel code generation, and the Matlab/Mathematica support for ClearSpeed processors. Basically, if what you have is a description on an algorithm rather than a particular implementation of it, compilation/code generatioin tools have much more information and freedom to make it run well on a particular piece of parallel hardware. 

Note that a domain-specific language can also be a kit of libraries and classes and algorithms used with a coding language (http://jakob.engbloms.se/archives/45 talks about one such example).</description>
		<content:encoded><![CDATA[<p>I agree that we need new languages, but I have a feeling that the successful ones will be those that are higher-level and more tailored to a domain. I talked to Jim on this in August this year, and since his background is mostly in HPC, he does not really have that kind of experience.  He thinks in &#8220;coding&#8221; languages like C/Fortran/Erlang. </p>
<p>I commented on this here: <a href="http://jakob.engbloms.se/archives/35#more-35" rel="nofollow">http://jakob.engbloms.se/archives/35#more-35</a> and here: <a href="http://jakob.engbloms.se/archives/55#more-55" rel="nofollow">http://jakob.engbloms.se/archives/55#more-55</a> . Anyone interested should check out what Labview is doing in parallel code generation, and the Matlab/Mathematica support for ClearSpeed processors. Basically, if what you have is a description on an algorithm rather than a particular implementation of it, compilation/code generatioin tools have much more information and freedom to make it run well on a particular piece of parallel hardware. </p>
<p>Note that a domain-specific language can also be a kit of libraries and classes and algorithms used with a coding language (http://jakob.engbloms.se/archives/45 talks about one such example).</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Stefan Eilemann</title>
		<link>http://www.thinkingparallel.com/2007/10/18/choice-overload-and-parallel-programming/#comment-27312</link>
		<dc:creator>Stefan Eilemann</dc:creator>
		<pubDate>Fri, 19 Oct 2007 09:53:45 +0000</pubDate>
		<guid isPermaLink="false">http://www.thinkingparallel.com/2007/10/18/choice-overload-and-parallel-programming/#comment-27312</guid>
		<description>I'm obviously excited about Equalizer, a parallel programming framework for OpenGL applications. It helps you to parallelize your OpenGL application to run on multi-GPU systems and graphics clusters, in order to scale the display size and/or rendering performance. Included are a polygonal and volume rendering example. Check out the website (link above) for more information.

Now back to work to get the v0.4 release out!


Cheers,

Stefan.</description>
		<content:encoded><![CDATA[<p>I&#8217;m obviously excited about Equalizer, a parallel programming framework for OpenGL applications. It helps you to parallelize your OpenGL application to run on multi-GPU systems and graphics clusters, in order to scale the display size and/or rendering performance. Included are a polygonal and volume rendering example. Check out the website (link above) for more information.</p>
<p>Now back to work to get the v0.4 release out!</p>
<p>Cheers,</p>
<p>Stefan.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: My Ghillie &#187; Choice Overload and Parallel Programming</title>
		<link>http://www.thinkingparallel.com/2007/10/18/choice-overload-and-parallel-programming/#comment-27213</link>
		<dc:creator>My Ghillie &#187; Choice Overload and Parallel Programming</dc:creator>
		<pubDate>Thu, 18 Oct 2007 19:13:03 +0000</pubDate>
		<guid isPermaLink="false">http://www.thinkingparallel.com/2007/10/18/choice-overload-and-parallel-programming/#comment-27213</guid>
		<description>[...] Check it out! While looking through the blogosphere we stumbled on an interesting post today.Here&#8217;s a quick excerptLet me compare the situation today with the search-engine market before Google&#8230;. [...]</description>
		<content:encoded><![CDATA[<p>[&#8230;] Check it out! While looking through the blogosphere we stumbled on an interesting post today.Here&#8217;s a quick excerptLet me compare the situation today with the search-engine market before Google&#8230;. [&#8230;]</p>
]]></content:encoded>
	</item>
</channel>
</rss>
