AMD's 3GHz K10 to break 30,000 3DMark06

@ 2007/08/29
LAST WEEK in Leipzig my kit was nicked, but before that happened we asked AMD if it would let us run memory benchmark scores on a system there. The reps gave us the company line and declined, so we decided to disclose the benchmark scores of our own K10 benchmarking here and now.

If you were wondering why AMD was hiding the scores of K10 so secretly, there were two reasons. The first might be that the CPU sucks badly and after AMD comes out, Intel's lads can start celebrating the death of AMD. On the other hand, there the was clear and present danger of the K10 significantly beating not just the current Conroe/Kentsfield generation, but easily out besting Wolfdale/Yorkfield. This statement warrants at least three hatemails from Intel's R&D lads, but all that we will disclose here are results we have in our possession. The pics are gone with my stolen laptop, though.

Comment from geoffrey @ 2007/08/30
My Croffire test on DFI RD600, C2D @ 3,23GHz -> 14579 3D Marks 2006 (http://www.madshrimps.be/?action=get...&articID= 581)

My other 'unofficial' test on P5B Del, CPU @ 3850MHz -> 15644 3D Marks 2006 (http://www.hwbot.org/compare.do?resultId=620731)

No VGA overclocking with both benchmarks.
No perfect compare, second run is more tweaked for high score, and boards are different too, but with ~ 600MHz higher clocked CPU I got around 1000 extra 3D Marks.

3D Mark 2006 does scale well when you overclock your CPU, though knowing that Conroe allready has a short pipeline I can hardly believe that there is so much more performance left.
Comment from Rutar @ 2007/08/30
Quote:
Originally Posted by jmke View Post
you compare C2D to C2Quad ?
Doesn't seem to have an influence on the GPU scores, which is interesting for my comparison as the bigger the CPU effect is on the GPU scores, the less it's worth.
Comment from jmke @ 2007/08/30
you compare C2D to C2Quad ?
Comment from Rutar @ 2007/08/30
And they still didn't calculate it the way I wanted too.


gotta do it myself ><

GPU score of a stock crossfire 2900 setup: 12980

http://service.futuremark.com/compare?3dm06=2694745

GPU score of a crossfire 2900 setup near INQ clocks: 13080

http://service.futuremark.com/compare?3dm06=2664231

so we get

CPU score stock: 10787

CPU score overclocked: 16951

A 57% increase. Does boosting the CPU clock increase the GPU scores that much that my results are completely off?
Comment from jmke @ 2007/08/30
knew OC.com would have their say

Quote:
First, the 3DMark score went up 26% just by increasing the processor speed 20%. That's a scaling of 130%. That's pretty good, matter of fact, it's too good to be true based solely on changing the speed of the same type CPU.

Second, this is especially true for 3DMark 2006, which is primarily a video card, not a CPU benchmark.
http://www.overclockers.com/tips01208/

3D06 is first 3DMark which takes into account CPU score though, the older benchmarks did not.
Comment from Jaco @ 2007/08/30
sounds too good to be true . I have my doubts...
Comment from Rutar @ 2007/08/29
Quote:
Originally Posted by jmke View Post
even with overclocked VGA, 30k is way over current top score.
that isn't the point, the point is that the CPU score would have to increase more than the clockspeed increase to reach this score
Comment from jmke @ 2007/08/29
Quote:
Originally Posted by Rutar View Post
Can someone run the numbers if the first Radeons were at stock and the second ones were overclocked, how much of an % increase of the CPU score has to be attributed to the 20% increase in CPU clock.


for me, the numbers don't add up since the boost seems bigger than the clockspeed boost
even with overclocked VGA, 30k is way over current top score.

@lazyman: Stolen story is true, car was plundered, camera, laptops, passports
http://www.theinquirer.net/default.aspx?article=41909
Comment from easypanic @ 2007/08/29
MMM, we need real numbers and facts...
Comment from Rutar @ 2007/08/29
Can someone run the numbers if the first Radeons were at stock and the second ones were overclocked, how much of an % increase of the CPU score has to be attributed to the 20% increase in CPU clock.


for me, the numbers don't add up since the boost seems bigger than the clockspeed boost
Comment from geoffrey @ 2007/08/29
I tried, but I just can not believe this.
Comment from Sidney @ 2007/08/29
Quote:
The pics are gone with my stolen laptop, though.
How convenient for a pro without a simple USB flash backup.
Comment from jmke @ 2007/08/29
Highest score for HD 2900 XT CF system on 3D06 at HWbot is here: http://www.hwbot.org/result.do?resultId=640671

22.224 with Core 2 QX6850 @ 4510mhz
2x Radeon HD 2900 XT @ 880/990mhz

they score with K10 Barcelona

30.031 with K10 @ 3000Mhz
2x Radeon HD 2900 XT @ 830/900mhz

~35% increase in score with SLOWER GPU and lower CPU clocks. K10 = the bomb?