CPU Benchmarking Project
This Show the level of Performance gained by compiling with the -O3 Flag on in
GCC.
Using the optimization in GCC give on average a 1.84 time performance improvement.
This improvement is most significant on the Pentium 4 CPU with a more and doubling of speed.
|
CPU |
Clock |
No Opt |
CC -O3 |
* improvement |
Server Name |
|
||
|
P4 |
2266 |
2.29 |
1.10 |
2.08 |
Nisvara |
FBSD 4.4 |
|
|
|
PIII Celeron |
533 |
6.72 |
3.75 |
1.79 |
1U Test |
FBSD 4.6.2 |
|
|
|
Xeon PII |
450 |
7.70 |
4.70 |
1.64 |
|
FBSD 3.0 |
|
|
|
AMD-K6 |
450 |
8.47 |
4.46 |
1.90 |
dnull |
FBSD 2.2.7 |
ELF Binary |
|
|
AMD-K6 |
450 |
8.50 |
5.25 |
1.62 |
dnull |
FBSD 2.2.7 |
|
|
|
AMD-K6 |
450 |
8.49 |
4.46 |
1.91 |
dnull |
FBSD 3.2 |
|
|
|
AMD-K6 |
450 |
9.12 |
4.63 |
1.97 |
Reactor-core |
Linux 2..4.19ac2 |
901.12 BogoMIPS |
|
|
Intel ? |
350 |
10.25 |
5.69 |
1.80 |
via.net |
FBSD 4.1.1 |
|
|
|
|
|
|
|
|
||||
|
|
|
|
1.84 |
Average performance improvement from using the GCC complier -O3 option. |
||||
Copyright © John L. Sokol
http://www.dnull.com/~sokol