View mode: basic / threaded / horizontal-split · Log in · Help
January 24, 2005
[performance]Win32.dmd against cygwin.dmd
Looks nice :-)

                     results
Program            Win32    cygin    unit   factor

t.ackermann(4,1)  29.547   13.985    s      2.11
dhrystone         3.079956 5.882353  DMIPS  1.91
whetstone.single  1048.670 1339.764  MWIPS  1.28
whetstone.double  1001.124 1284.687  MWIPS  1.28
whetstone.real     716.854  881.343  MWIPS  1.23

-manfred
January 29, 2005
Re: [performance]Win32.dmd against cygwin.dmd
In article <ct3hp2$1upg$1@digitaldaemon.com>, Manfred Nowak says...
>
>Looks nice :-)
>
>                      results
>Program            Win32    cygin    unit   factor
>
>t.ackermann(4,1)  29.547   13.985    s      2.11
>dhrystone         3.079956 5.882353  DMIPS  1.91
>whetstone.single  1048.670 1339.764  MWIPS  1.28
>whetstone.double  1001.124 1284.687  MWIPS  1.28
>whetstone.real     716.854  881.343  MWIPS  1.23
>
>-manfred

I'm confused is this DMD-Win32 (Digital Mars compiler) versus GDC using Cygwin
on Windows?

Did you optimize the DMD stuff?
January 30, 2005
Re: [performance]Win32.dmd against cygwin.dmd
Mark T wrote: 

> I'm confused is this DMD-Win32 (Digital Mars compiler) versus GDC
> using Cygwin on Windows?
> 
> Did you optimize the DMD stuff?

Yes. The programs were compiled unmodified with Digital Mars compiler 
dmd.0.111.win32 and gdc-GCC3.4.4 20050121-prerelease.cygwin both 
compiles and resulting executabls ran under WinXP.PRO.servicepack2.

The dmd.win32-compile and the gdc.cygwin.win32.compile using the dmd 
script used the flags "-O -release -inline".

The programs compiled are the ackermann program from D-Stress 
benchmarks, the dhrystone program from the examples coming with dmd and  
my adaption of the whetstone benchamrk to D, for which I do not know, 
whether I am allowed to publish it.

-manfred
Top | Discussion index | About this forum | D home