Hi Tim,

 Thanks for your comments. I think I should use
time as you say. I will try it out. The reason I measure
outside the code is that I would like to include compilation
time also. In fact the gcc tests I run like

date;gcc -O5 -o mandel mandel.c -lm;./mandel;date

This is to make the comparison more "even" with other
non-compiled code (PDL, MatLab). In the end it is not
a big difference since the time to compile is much less
than the time to run.

MatLab starts pretty fast on my machine (without windows that
is, i.e., matlab -nodisplay -nojvm).
Just a few seconds.
So, again, the "compilation" time is not affecting significantly.

I am not trying to make a really serious benchmark test here.
I just want to make sure I am not making any huge mistakes.

 Xavier


On 3/1/07, Timothy Brown <[EMAIL PROTECTED]> wrote:
On Thu, Mar 01, 2007 at 07:32:42AM +0000, Xavier Calbet wrote:
[SNIP]
> I have run these benchmarks without running anything else on the computer.
> I have done this several times and have obtained consistent times.
> Is there a better way to measure the time without resorting to date?
>
> The results are the following:
>
> gcc             201 sec
> g77             201 sec
> PDL            651 sec
> IDL             694 sec
> MatLab     2738 sec
> Octave      2031 sec

I'll have a look at the code in a bit, but just wanted to say, I'd use
`time` instead of `date`. As this gives you the amount of time spent in
userland, system overhead and execution. I prefer the break down
really. (Usage would be: `time ./mandel.pl` )

Also do you care about startup times? As in Matlab takes a good half
minute or there abouts to start up on my machine... Would it make more
sense to do the timing within the languages? Then you could go further
and iterate the function. I don't know, how much detail do you really
want to get into?

Regards
Timothy


_______________________________________________
Perldl mailing list
[email protected]
http://mailman.jach.hawaii.edu/mailman/listinfo/perldl


_______________________________________________
Perldl mailing list
[email protected]
http://mailman.jach.hawaii.edu/mailman/listinfo/perldl

Reply via email to