Re: Similar systems, different performance of fortran code

[Victor Munoz <vmunoz@macul.ciencias.uchile.cl>]

> 
> of course, a better place to start would be to use same compilers on
> each machine and then, maybe, you can rule out the compiler version as a
> cause... 
> 

Ok, some 'time' results for this problem (file-<version> means compiled with
g77-<version>: g77-<version> -o file-<version> file.f

System A (theoretically slower, Pentium 4 2.40GHz, woody)

time file-2.95 

real 0m24.887s
user 0m20.530s
sys  0m3.390s

time file-3.0

real 0m18.528s
user 0m15.410s
sys  0m2.470s

System B (theoretically faster, Pentium 4 3GHz, sid)

time file-2.95 

real 0m15.576s
user 0m12.560s
sys  0m2.730s

time file-3.3

real 1m49.923s
user 1m47.330s
sys  0m2.430s

time file-3.4

real 0m13.077s
user 0m9.630s
sys  0m3.260s

Only 2.95 is available from debs for both woody and sid. Anyway, it is clear
there is some specific problem with g77-3.3, as both 2.95 and 3.4 in sid
perform better than 2.95 and 3.0 in woody.

Still investigating...

Regards,
					Victor