[Mono-list] Why is mono so slow at math compared to .Net?

[Rodrigo Kumpera]

Well, given the huge amount of R&D MS has put on .NET this doesn't come to a
surprise.

If code is 7x slower this is a performance bug on mono's side. This kind of
artifact is possible.

Mono weaker point is the garbage collector. If your code churns tons of
objects, the performance
figures are worse.

Are you running on 32 or 64 bits? Intel x86 or itanium? Just post the result
of "mono --version", please.



On Wed, Feb 18, 2009 at 5:30 PM, Robert Weiss <rweiss at chesapeakebay.net>wrote:

>  No **more** than 50% slower??? I really hope you are kidding me.
>
>
>
> Original App:
>
> C# .Net
>
> .Net Framework 2.0
>
> Windows XP Pro
>
>
>
> Ported Platform:
>
> Red Hat Enterprise Linux 5.3
>
> Mono 2.2
>
> C# .Net
>
>
>
> Tester coming shortly.
>  ------------------------------
>
> *From:* Rodrigo Kumpera [mailto:kumpera at gmail.com]
> *Sent:* Wednesday, February 18, 2009 3:26 PM
> *To:* Robert Weiss
> *Cc:* mono-list at lists.ximian.com
> *Subject:* Re: [Mono-list] Why is mono so slow at math compared to .Net?
>
>
>
>
>
> On Wed, Feb 18, 2009 at 5:17 PM, Rob Weiss <rweiss at chesapeakebay.net>
> wrote:
>
>
> We have ported a scientific app to RHEL-mono. It runs 7 times slower than
> it
> does in native .Net on windows.
>
> It does tons floating point math and array processes.
>
> Thoughts? Fixes?
>
>
> Which version of mono and archtecture you ran it?
>
> Mono should be no more than 50% slower that .NET on FP intensive code.
> Please
> fill a bug report with a test for your issue.
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.ximian.com/pipermail/mono-list/attachments/20090218/f30bc605/attachment.html