c***r
发帖数: 4631 | 1 it looks that we are talking different stuff.
People like me like to use "gcc -O3" or "gcc -funroll-all-loops" to get better
performance of program. I do numerical computation on my notebook either. And
by this way, the compiled program will become larger, and the "foot print"
will become larger. So this method may NOT get quicker program. If the cache
is large enough, performance will become better. If the cache is not large
enough, the program will run S L O W E R th |
|
