<?xml version="1.0" encoding="UTF-8" ?>
<?xml-stylesheet type="text/xsl" href="http://blogs.msdn.com/utility/FeedStylesheets/rss.xsl" media="screen"?><rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:slash="http://purl.org/rss/1.0/modules/slash/" xmlns:wfw="http://wellformedweb.org/CommentAPI/"><channel><title>CLR Performance Questionnaire</title><link>http://blogs.msdn.com/ricom/archive/2009/06/15/clr-performance-questionnaire.aspx</link><description>Vance Morrison posted an interesting opportunity for you to give feedback directly to the CLR perf team on your managed code performance issues. I encourage you to participate: If you use managed code, and you care about performance, then we want to hear</description><dc:language>en-US</dc:language><generator>CommunityServer 2.1 SP1 (Build: 61025.2)</generator><item><title>re: CLR Performance Questionnaire</title><link>http://blogs.msdn.com/ricom/archive/2009/06/15/clr-performance-questionnaire.aspx#9757842</link><pubDate>Tue, 16 Jun 2009 07:10:03 GMT</pubDate><guid isPermaLink="false">91d46819-8472-40ad-a661-2c78acb4018c:9757842</guid><dc:creator>cwbrandsma</dc:creator><description>&lt;p&gt;*cough* .netcf *cough*&lt;/p&gt;
&lt;p&gt;Sorry about that, something was caught in my throat. &amp;nbsp;Now what was that?&lt;/p&gt;
</description></item><item><title>re: CLR Performance Questionnaire</title><link>http://blogs.msdn.com/ricom/archive/2009/06/15/clr-performance-questionnaire.aspx#9762710</link><pubDate>Tue, 16 Jun 2009 20:24:43 GMT</pubDate><guid isPermaLink="false">91d46819-8472-40ad-a661-2c78acb4018c:9762710</guid><dc:creator>John Oglesby</dc:creator><description>&lt;p&gt;We write high level financial analytics in .NET languages. Further down the code stack we use C++, in the main for performance reasons. C++ is more difficult for to write and debug, so we have a big interest in the performance improving. We would like to us F#/C# for most of our code but appears too slow today.&lt;/p&gt;
&lt;p&gt;Our code is mostly in memory numeric manipulation, arrays of double, some trig functions etc.&lt;/p&gt;
&lt;p&gt;John&lt;/p&gt;
</description></item><item><title>re: CLR Performance Questionnaire</title><link>http://blogs.msdn.com/ricom/archive/2009/06/15/clr-performance-questionnaire.aspx#9763347</link><pubDate>Wed, 17 Jun 2009 02:06:51 GMT</pubDate><guid isPermaLink="false">91d46819-8472-40ad-a661-2c78acb4018c:9763347</guid><dc:creator>The Fiddler</dc:creator><description>&lt;p&gt;We mainly work with 3d graphics and physics simulation. The Mono runtime has a significant advantage in this area, due to its faster p/invoke implementation and the Mono.Simd library, which exposes SIMD CPU extensions to managed code (needs runtime support, obviously).&lt;/p&gt;
&lt;p&gt;Using Mono.Simd, we have seen performance improvements in the region of 2x-10x - which is very significant in our field of work!&lt;/p&gt;
</description></item><item><title>re: CLR Performance Questionnaire</title><link>http://blogs.msdn.com/ricom/archive/2009/06/15/clr-performance-questionnaire.aspx#9769828</link><pubDate>Wed, 17 Jun 2009 19:53:03 GMT</pubDate><guid isPermaLink="false">91d46819-8472-40ad-a661-2c78acb4018c:9769828</guid><dc:creator>ShuggyCoUk</dc:creator><description>&lt;p&gt;++ same as John&lt;/p&gt;
&lt;p&gt;We use mix of c#, f#, C++/CLI&lt;/p&gt;
&lt;p&gt;C++ where absolutely needed (often efficient interop with other c/c++ libraries)&lt;/p&gt;
&lt;p&gt;I filled in the questionnaire already :)&lt;/p&gt;
</description></item><item><title>re: CLR Performance Questionnaire</title><link>http://blogs.msdn.com/ricom/archive/2009/06/15/clr-performance-questionnaire.aspx#9792166</link><pubDate>Sat, 20 Jun 2009 00:27:25 GMT</pubDate><guid isPermaLink="false">91d46819-8472-40ad-a661-2c78acb4018c:9792166</guid><dc:creator>Keith Patrick</dc:creator><description>&lt;p&gt;Can you suggest to them that &amp;quot;Please specify the version of the .NET Framework that you are using?&amp;quot; should be a list of checkboxes rather than radio? &amp;nbsp;Some of us aren't fortunate enough to have our entire org shift BCL version across all apps simultaneously (I currenly use the last 3 versions and every now and then even have to code to 1.1)&lt;/p&gt;
</description></item><item><title>re: CLR Performance Questionnaire</title><link>http://blogs.msdn.com/ricom/archive/2009/06/15/clr-performance-questionnaire.aspx#9798931</link><pubDate>Tue, 23 Jun 2009 09:00:35 GMT</pubDate><guid isPermaLink="false">91d46819-8472-40ad-a661-2c78acb4018c:9798931</guid><dc:creator>Justin_H</dc:creator><description>&lt;p&gt;I also ended up having to just check .NET 2.0 in the questionnaire even though I also work on 3.5 and 1.1. Should have been check boxes I agree. &lt;/p&gt;
&lt;p&gt;Thanks Fiddler for the info on Mono. I never would have considered using Mono on windows or that it may perform better in some scenarios. &lt;/p&gt;
</description></item><item><title>re: CLR Performance Questionnaire</title><link>http://blogs.msdn.com/ricom/archive/2009/06/15/clr-performance-questionnaire.aspx#9818415</link><pubDate>Mon, 06 Jul 2009 04:15:09 GMT</pubDate><guid isPermaLink="false">91d46819-8472-40ad-a661-2c78acb4018c:9818415</guid><dc:creator>Julian</dc:creator><description>&lt;p&gt;It would indeed be awesome if .NET had something similar to Mono.Simd. And while I've heard good things about Mono.Simd, I still think it's kind of weird to have a small set of structs that are treated very differently by the runtime than all the rest. Doesn't feel right. I suppose it would be more '.NET' if you could decorate a ValueType with an attribute, but I have no idea on the technical feasibility of that. &lt;/p&gt;
</description></item><item><title>re: CLR Performance Questionnaire</title><link>http://blogs.msdn.com/ricom/archive/2009/06/15/clr-performance-questionnaire.aspx#9826244</link><pubDate>Thu, 09 Jul 2009 12:53:09 GMT</pubDate><guid isPermaLink="false">91d46819-8472-40ad-a661-2c78acb4018c:9826244</guid><dc:creator>Pop.Catalin</dc:creator><description>&lt;p&gt;What I'd love to see in a future .Net runtime is:&lt;/p&gt;
&lt;p&gt; - better am more efficient string manipulation (lot's of room for perf and memory optimizations here)&lt;/p&gt;
&lt;p&gt; - inlining of delegate calls (at least those that are created in scope)&lt;/p&gt;
&lt;p&gt; - SIMD support (Hardware agnostic SIMD would be really nice :) )&lt;/p&gt;
&lt;p&gt;There are areas in computing where .Net is a bit out of reach like, numerics, simulations, 3D graphics, and it's a real shame for such a beautiful runtime to be left out ... but, I'm sure one day we will get there; you guys have the gift of exceeding my expectations when you finally add the things I wished for.&lt;/p&gt;
</description></item></channel></rss>