Cezary Czaplewski writes:
>
>
>I have found gcc/g77 2.95.2 faster than egcs-1.1.2 , I think mostly
>because of -march=pentiumpro optimization, but I didn't do extensive
>testing. For some our codes I have found that commercial PGF90 is superior
>to g77 2.95.2 but for SANDER difference was small.
I just checked by compiling AMBER with gcc/g77 2.95.2 and egcs-1.1.2. 2.95.2
is about 15-20% faster than egcs-1.1.2 for the exact same switch.
I saw no observerable difference in speed between -O3 and -O3 -march=pentiumpro,
however. I still haven't found the "perfect" set of optimization flags or
compiler version yet.
Dave
Received on Wed Jun 14 2000 - 20:53:54 PDT
