Rockfella said:
It wont make much difference in the games u playing @ that reso. Chuck OCing.
dOm1naTOr said:
res is the culprit. Sam 2233 needs 1920*1080. At that res, the GPU is the one which is bottlenecked. CPU clcoks hardly matters.
To convince him, run 3dm 06 on stock and oced clocks. Surelly ull get bout 1k CPU score boost and bout 2~2.5k overall score boost if u reach atleast 3.3Ghz.
That isn't a rule engraved in stone. You cannot ignore the game in question and its CPU:GPU bias ratio. More often then not this bias shifts to make up for the weaker links in a system. So if you're talking about high resolutions the CPU will make up for a weaker GPU if it can, and vice versa(+1 to dOm1naTOr's last post). As for lower resolutions:
To put things into perspective i just did a small test:-
Difference in frame rates(average) between E5200 @ stock (2.5 Ghz, 800Mhz FSB) and Overclocked (3.3Ghz, 1200Mhz FSB):
Prototype:-
Before: 30 FPS. , After: 45 FPS
Tom Clancy's HAWX(DX9):-
Before: 120 FPS. , After: 180 FPS
RaceDriver GRID:-
Before: 45 FPS. , After: 55 FPS
Street Fighter IV:-
Before: 55-60 FPS, After: 60 FPS (Constant) [Think the game forces vsync, need to confirm)
Just to make things clear, all of this is at a resolution of
1024x768, to put the "at low res. CPU offloads all 3D processing to the GPU" out the window.
All eye candy on/maxed.
As you can see, all the above games show improvements, some more than others for obvious reasons, for eg, Prototype and HAWX, both CPU intensive, whereas SFIV shows a very small increase.
To elaborate further, ran 3D mark06 at the same resolution:-
Before: 10206
After: 12196
A 2k score improvement is huge comparatively, as expected the real world performance doesn't show such an improvment(comparatively), but you can in now way discount its effect. This isn't a query to which you can get an answer going by the book since every game engine utilizes resources differently.
This is the reason i asked OP to run multiple games.
Hope this helps clear any doubts.
