Andy Ross wrote:

On my laptop (a 2.2GHz P4 that is currently reporting 1196MHz in
/proc/cpuinfo), compiled with -O2 (no fancy optimizer settings), I'm
seeing the following:

localhost:~$ time ./geod andy
1000000 iterations

real    0m4.097s
user    0m4.060s
sys     0m0.020s

localhost:~$ time ./geod toms
1000000 iterations

real    0m0.871s
user    0m0.870s
sys     0m0.000s

That comes out to about 4900 CPU cycles per call for my
implementation, and 1042 cycles for the GEODETIC code.  Both are
fairly small numbers for a function called only a handful of times
each frame.

Just for comparison, my results on an O2/300 Mhz RM5200 CPU, compiled with MipsPro cc and -O2:


time ./geod-perf andy
1000000 iterations

real    0m8.483s
user    0m8.133s
sys     0m0.027s

time ./geod-perf toms
1000000 iterations

real    0m2.794s
user    0m2.605s
sys     0m0.015s

Erik


_______________________________________________ Flightgear-devel mailing list [EMAIL PROTECTED] http://mail.flightgear.org/mailman/listinfo/flightgear-devel

Reply via email to