summing large arrays - or - performance of tight loops

[Manfred Nowak]

Dave wrote
> Mine were ran on a P4 chip and an AMD64 chip (original was
> Turion). 

It is disgusting to see, that three random benchmarkers all get 
different results. Mine were run on an AMD64 X2 under WinXP32 and 
compiled with DMD1.012.

Quotients (foreach/other):
Thomas: 200%
Dave:   100%
me:      85%

What causes those huge differences?

Here is my main:

import std.stdio, std.perf;
void main(){
  const int f= 256;
  int[ ] a= new int[ 1024*1024*f];
  auto c=new PerformanceCounter;
  auto c2=new PerformanceCounter;
  c.start;
  for(int i= 1; i<=5120/f; i++)
    volatile sum_foreach( a);
  c.stop;
  volatile auto t= c.microseconds;
  
  c2.start;
  for(int i= 1; i<=5120/f; i++)
    volatile sum_array( a);
  c2.stop;
  writefln( "%s %s %s"
    , cast(real)t/c2.microseconds
    , t
    , c2.microseconds
  );
}

-manfred