[Lemon-devel] Strange running time results again

[Alpár Jüttner]

Hi,

I managed attached the source instead of the results in my last mail.
Sorry for the confusion, I send the right attachment now.


On Wed, 2007-06-27 at 10:27 +0200, Alpár Jüttner wrote:
> Hi,
> 
> I compiled the code you can see below. I used g++-4.1.2 with
> optimization flag -O2. The attached picture shows the running times of
> the iterations.
> 
> Could anyone explain that strange "forking" tendency in the running
> times?
> 
> Regards,
> Alpar
> 
> #include <lemon/time_measure.h>
> #include <iostream>
> 
> double mpower_sum(int size,double *m)
> {
>   double s=0;
>   for(int i=0;i<size;i++)
>     for(int j=0;j<size;j++)
>       for(int k=0;k<size;k++)
> 	s+=m[i*size+k]*m[k*size+j];
>   return s;
> }
> 
> int main(int argc,char **argv)
> {
>   int size=1000;
>   int step=10;
>   
>   double *m = (double *)malloc(sizeof(double)*(size*size+3));
>   for(int i=0;i<size*size;i++) m[i]=i;
>   for(int i=step;i<=size;i+=step)
>     {
>       lemon::Timer t;
>       mpower_sum(i,m);
>       std::cout << i << ' ' << t.realTime() << std::endl;
>     }
>   free(p);
> }
> 
> _______________________________________________
> Lemon-devel mailing list
> Lemon-devel at lemon.cs.elte.hu
> http://lemon.cs.elte.hu/mailman/listinfo/lemon-devel
-------------- next part --------------
A non-text attachment was scrubbed...
Name: mpower.png
Type: image/png
Size: 4710 bytes
Desc: not available
URL: <http://lemon.cs.elte.hu/pipermail/lemon-devel/attachments/20070627/8b64788b/attachment.png>