Official 3DMark 11 Thread

P9903 3DMarks
http://3dmark.com/3dm11/1210314
3DMark Score P9903
Graphics Score 9988
Physics Score 10918
Combined Score 8231
GraphicsTest1 43.31 FPS
GraphicsTest2 53.06 FPS
GraphicsTest3 63.85 FPS
GraphicsTest4 28.98 FPS
PhysicsTest 34.66 FPS
CombinedTest 38.28 FPS

fresh install of 11.5a

[Moshing]

Just as an interesting observation if I may, it looks like generally, in 3Dmark11, there's a pattern for scoring that can be somewhat followed.

As long as one is using any higher end, 4Ghz+ Intel Core i7 (first or second gen), then one can follow this pattern pretty closely.

Every GTX 580 added is worth ~5000 points.
Every HD6970 added is worth ~4500 points.

Both seem to scale nearly linearly, so add 2 GTX 580 = ~10k, while 2 6970 fall in the 9K range. 3 GTX 580 =15K, and so on. Excessive overclocking can push scores up to nearly 20K with 3 GTX 580's and a crazy clocked i7.

[Roadhog]

Originally Posted by Moshing

Just as an interesting observation if I may, it looks like generally, in 3Dmark11, there's a pattern for scoring that can be somewhat followed.

As long as one is using any higher end, 4Ghz+ Intel Core i7 (first or second gen), then one can follow this pattern pretty closely.

Every GTX 580 added is worth ~5000 points.
Every HD6970 added is worth ~4500 points.

Both seem to scale nearly linearly, so add 2 GTX 580 = ~10k, while 2 6970 fall in the 9K range. 3 GTX 580 =15K, and so on. Excessive overclocking can push scores up to nearly 20K with 3 GTX 580's and a crazy clocked i7.

Are you talking about total score or just graphics score? Because the total score can vary widely due to faster cpus in the physics benchmarks.

[Moshing]

Originally Posted by Roadhog

Are you talking about total score or just graphics score? Because the total score can vary widely due to faster cpus in the physics benchmarks.

It certainly can! But at around 4Ghz (4-4.2ish?) on first or second gen i7, as my example stated, that does seem to be the averages. Over 4Ghz pushes that Physics score higher, I get around 11K on the physics score myself with 5Ghz of second gen i7. And this is only averages, but provides a handy rule-of-thumb that certainly has exceptions, but not many within the scope of description.

@ 4Ghz
P9994 3DMarks
http://3dmark.com/3dm11/1212853

[shadow001]

Another planet burner reporting in....Still in descovery of what this setup can do though..


http://3dmark.com/3dm11/1220414;jses...pLAFMCbCpbmNqn


P16194

[Roadhog]

Originally Posted by shadow001

Another planet burner reporting in....Still in descovery of what this setup can do though..


http://3dmark.com/3dm11/1220414;jses...pLAFMCbCpbmNqn


P16194

Something is seriously wrong with your setup. I get higher fps than you in one test, and the other ones are barely higher.

http://3dmark.com/3dm11/1006729

[Moshing]

Originally Posted by Roadhog

Something is seriously wrong with your setup.

Yeah, it's a pure drag racer with little real world usage and loses in PCMark 7 brutally to machines 1/3rd the cost

[shadow001]

Originally Posted by Roadhog

Something is seriously wrong with your setup. I get higher fps than you in one test, and the other ones are barely higher.

http://3dmark.com/3dm11/1006729

The only one where it's really close is graphics test 2(2 Fps difference), the other 3 graphics tests are 7 to 12 Fps higher average for me....The physics run is 15 Fps higher for me too ,and the combined test is 12 Fps higher as well...


kinda loose interpretation of barely higher isn't , and your CPU is running 500 Mhz higher than my pair to boot, and only the physics and combined tests are using all 12 cores in my setup....The first 4 test don't...

[shadow001]

Originally Posted by Moshing

Yeah, it's a pure drag racer with little real world usage and loses in PCMark 7 brutally to machines 1/3rd the cost

Funny, i just played a few rounds of bad company with the drag racer running at the exact same clocks it did in that 3Dmark run....Yes, it can be faster still.

[Roadhog]

Originally Posted by shadow001

The only one where it's really close is graphics test 2(2 Fps difference), the other 3 graphics tests are 7 to 12 Fps higher average for me....The physics run is 15 Fps higher for me too ,and the combined test is 12 Fps higher as well...


kinda loose interpretation of barely higher isn't , and your CPU is running 500 Mhz higher than my pair to boot, and only the physics and combined tests are using all 12 cores in my setup....The first 4 test don't...

Are you sure the memory isn't overclocked too high? Your score should still be higher... You will never know if you go too high on DDR5 as it has ECC. It will just hurt performance and still be stable.

EDIT: If it makes you happy here is my 4ghz run. Same exact fps so you can't pull the clock speed card this time.

http://3dmark.com/3dm11/914071

[Moshing]

Originally Posted by shadow001

Funny, i just played a few rounds of bad company with the drag racer running at the exact same clocks it did in that 3Dmark run....Yes, it can be faster still.

Indeed it can. 4Ghz isn't really pushing anything very hard on the CPUs is it? Anything less then 4.6Ghz on that machine will kinda be disappointing, eh? You can make up the score difference, I'm sure of it. Just PUSH those procs already, you're not even making them sweat!