Watercooled Phenom II X4 B55 Overclocked gives lower scores.

Dags

Honorable
Aug 5, 2013
141
0
10,710
Hi everyone.
I have a Phenom II X2 555 BE with all 4 cores unlocked and currently running at 4.0GHz super stable with 1.45v (which I believe it is a bit too much) on a Gigabyte GA-870A-UD3 Watercooled by the old Corsair HydroCool 200 http://www.modthebox.com/review264_1.shtml with it's tubes replaced by 1/2" and waterblocked replaced by this awesome Greek OCLabs monoblock http://www.oclabs.com.
Idle temperature (on a summer night) 26C and With Prime95 running for like 20min 44C.
Problem is that I had 7.7 CPU WEI with 4 cores at 3.4GHz and now after OC at 4GHzI have 7.5. I got to 4GHz by only raising the multiplier and voltage. Am I doing anything wrong?
The other day I heard an Intel employee saying that sometimes its better to lower the internal clock and raise multiplier or the opposite. Is that true? Can that be the reason?
 
Don't worry about WEI, Synthetic scores like that are good for a baseline dirty rating on an upgrade at best.
To test your performance and find out how any Overclock or upgrade has benefited you/or not its best practice to benchmark with real world usage tests.
If you are a gamer then run a few benchmarks on actual games. If you are looking for an encoding performance boost then check out the difference in encoding times after any changes.

Mactronix :)
 

Dags

Honorable
Aug 5, 2013
141
0
10,710
Thanks for your reply. I am no longer a gamer. But what I do care about and the reason I watercooled-overclocked is performance on productivity-encoding. Instead of upgrading to an FX cpu (waiting for steamroller) I threw away the Air Cooler and replaced it with my old good HydroCool which I haven't used for years. Even waterblock bracket (once on a K7 socket) was compatible. So I am trying to get as much juice out as possible without spending a penny until I upgrade to a new CPU or Platform.
Do you know any benchmark suites that do not measure graphics?