Kentsfield E6700 1M U10

[sierra_bound]

Very nice result, kaz-n-san.

[Rol-Co]

now that's a great score on cinebench =D
Great to see that P5B can give a much higher fsb with their i965 then i975x one's
is this the perfect combo (asus P5B+kentsfield)?

p5b is the only board atm who has the multi working right.......so higher fsb possible.
450 fsb done with kentsfield

[kaz-n]

Thanks TaPaKaH-san,charlie-san,Sesto Sento-san,sierra_bound-san,Rol-Co-san.
BIOS 0614 FSB limit B0 is 440, B1 is 420.
I will wait Next P5B BIOS or P5W DH,64 New BIOS.

[***Deimos***]

C O N G R A T U L A T I O N S[/color]

[Punisher!]

Amazing! :d

[ANP !!!]

Woowoowoow

I m Speechless.

[rob725]

@sierra:
You predicted better cinebench with lower clocks/higher fsb. Which is exactly what happened: >5% less cpu clock produced about 15% higher score running on a chipset that is usually slower than 975. Sorry for the newb question, but would you mind explaining? Just better ram speed?

[sierra_bound]

Cinebench is a very FSB-dependant bench. CPU speed in and of itself won't necessarily give you a good score.

If you can get FSB above 400, you have a good shot of scoring over 2000, even with air cooling.

Here's an example. I did this awhile ago with air cooling. Only needed 3.7GHz to score above 2000. Conroe won't get anywhere close to this score, regardless of cooling.

[hipro5]

Nice work kaz-n......Keep it going.....

[rob725]

Thanks. So, do you suppose that's due to increased fsb = faster inter-core communication?

[sierra_bound]

I'm not sure of the technical reason. I just know this from personal experience. FSB makes a bigger different than CPU or RAM speed. Notice how kaz-n's score jumped after he switched to the P5B Deluxe and ran 400+ FSB. The P5B Dlx is the only board at the moment that can do 400+ FSB with Kentsfield.

[rob725]

Have you noticed the same thing (faster fsb being more important) when doing video rendering with kentsfield, or is the difference to subtle to see outside of a benchmark?

[sierra_bound]

Did not really noticed it in other things. But after running Cinebench a number of times, I began to notice that CPU speed by itself wasn't the key to getting a good score. BTW, my Kentsfield has now been passed on to a very capable overclocker.

[Rol-Co]

bandwidth is important, but there is one problem with kentsfield and the boards

p5b is the only board who got the multiplier right......rest of the asus boards aren't
p5b + 0614 bios
http://82.173.172.10/Rol-Co/445.png

asus p5w dh i'm getting this, so multi is no use

http://82.173.172.10/Rol-Co/multi2.png

but at lower fsb and higher cpu speed you can make a nice score.

http://82.173.172.10/Rol-Co/cine3730.png

all at air cooling

sorry kaz-N san for going a little oftopic.
how do you get the multi 12 working on the p5b , i got the same bios but the step 4 b0 doesnt do that.

[kaz-n]

Hi,Rol-Co-san.
In BIOS, a magnification variable was set to ENABLE and it was able to carry out variable ordinarily. I think that it was possible to 20 or more times surely.