Concerns about performance w/Python, Pysco on Pentiums

[Peter Hansen]

Chris Liechti wrote:
> 
> Peter Hansen <peter at engcorp.com> wrote:
> > More interesting to me, however, is the poor relative performance
> > of the faster machines.  
> 
> the P4 does not seem te be equaly efficient with it cycles as a P3 ;-)
> 
> see e..g here, second diagramm and the text below:
> http://www.sysopt.com/reviews/pentium4/index4.html

Thanks for that link!  I did spend some time googling but didn't
manage to find anything that showed things that clearly.  The pages
I found vaguely gave the impression I should expect a 100% boost,
not just 25-30%.  That explains part of the problem (but not the Psyco
part).

> in addidtion to that different compilers (and options) might explain a
> bigger difference. i assume that you used the same python versions and no
> other CPU hogging processes...

Same versions (2.2) and no load other than the test, definitely.  I
think you hit the nail on the head with this one, as the test appears
to me (in spite of Gerhard's good suggestion) to be very nearly *just*
raw CPU usage.

I'll profile a bit to learn more...  but does anyone have any
comments on the Psyco front?  

-Peter