| 1 |
Hello, I've just upgraded my -rc6 to -1.0-gcc3 and decided to make an |
| 2 |
(unofficial) benchmark. |
| 3 |
|
| 4 |
I went for galeon, I had originally intended to use mozilla, but the |
| 5 |
time-results borked so I go for galeon instead.. smaller codebase, so |
| 6 |
its not as great difference, but it does have both c and c++ code, so it |
| 7 |
might be a decent choice. |
| 8 |
|
| 9 |
|
| 10 |
gcc 2.95.3 : |
| 11 |
real 3m38.592s |
| 12 |
user 2m46.810s |
| 13 |
sys 0m28.100s |
| 14 |
CFLAGS="-march=i686 -O3 -pipe" |
| 15 |
CXXFLAGS="-march=i686 -O3 -pipe" |
| 16 |
|
| 17 |
|
| 18 |
gcc 3.0.4 : |
| 19 |
real 5m6.465s |
| 20 |
user 3m27.440s |
| 21 |
sys 0m30.140s |
| 22 |
CFLAGS="-march=athlon -O3 -pipe" |
| 23 |
CXXFLAGS="-march=athlon -O3 -pipe" |
| 24 |
|
| 25 |
|
| 26 |
|
| 27 |
|
| 28 |
if you only compare the "user" time it should be enough... as the "sys" |
| 29 |
show, there's a few percentages difference between them, so this is not |
| 30 |
scientific or anything. |
| 31 |
|
| 32 |
Would be interesting to compare the results as well, since those are |
| 33 |
quite likely rather different with the new levels of optimization... |
| 34 |
|
| 35 |
|
| 36 |
//Spider |
| 37 |
|
| 38 |
|
| 39 |
-- |
| 40 |
begin happy99.exe |
| 41 |
This is a .signature virus! Please copy me into your .signature! |
| 42 |
See Microsoft KB Article Q265230 for more information. |
| 43 |
end |
Archives