On Fri, 5 Jul 2002 20:19:30 +1000, Nick Wienholt <[EMAIL PROTECTED]>
wrote:
>I am seeing something very different from your results.
>
>1. An initial capacity is much better for performance. (x3 better for no
>boxing)
That is odd. It is most definitely a wash on my system (even a few ms
slower) in both C# and C++.
Anyway, my comment about preallocation was directed toward the cases of
dynamically allocated objects (case C & D). I consider the boxing case to
be more interesting since, in an objected oriented design you will not
generally have arrays of basic types (int, double).
I think the point is that in general the allocations for all the objects
within the array will is the overwhelming factor.
>2. Not too much difference between builds, but release definitely better.
Ops, I wasn't clear that Debug and Release are for C#.
C# builds
____________/
/ \
> > | Debug Release Release
> > Boxing Initial Capacity | ms ms C++ ms
> > ------ ---------------- | ----- ------- -------
> > A No Yes | 300 260 1061
> > B No No | 1271 1281 981
> > C Yes Yes | 7590 7701 3144
> > D Yes No | 6889 7010 2954
I just found it mildly interesting that there was not a bigger difference
between debug and release builds.
Cheers...David
You can read messages from the Advanced DOTNET archive, unsubscribe from Advanced
DOTNET, or
subscribe to other DevelopMentor lists at http://discuss.develop.com.
