is this 3dmark score any good

[flymo]

just overclocked my GPU to GTX speeds, ran 3dmark 05 and got this result



is this a good score??

john

[NicksFXHouse]

set your drivers up like this and run it again....

NVIDIA DRIVER SETTINGS

AF=16x
Anisotropic Mip filter optimization = ON
Anisotropic sample optimization = ON
AA = 4x
Conformant texture clamp = use hardware
Extension limit = OFF
Force MipMaps = none
Gamma correct AA = ON
Hardware Acceleration = Single Display Performance mode
Ignore open GL error = ON
Negative LOD BIOS = CLAMP
OpenGL threading = AUTO
Texture Filtering= HIGH QUALITY
Transparency AA = OFF
Trillinear Optimizationfiltered=ON
Triple Buffering = ON
Vertical Sync = use 3D Application

It is important for LCD the scaling in the Nivida control panel is set to:

'Use my displays built in scaling" or "Monitor Scaling"


also... make sure you are running the default test and do not change any of the 3DMark options, just click start


if you get the same score or higher after that I would say your overclock is very good.

[ctjoyce]

just overclocked my GPU to GTX speeds, ran 3dmark 05 and got this result



is this a good score??

john

I hate you. The best I can pull off is in the low 9,000s

But use Nicks advice to find out your optiomal.

Cheers
Cameron

[NicksFXHouse]

I hate you. The best I can pull off is in the low 9,000s

But use Nicks advice to find out your optiomal.

Cheers
Cameron

If I can get 8400 out of an AGP rig with the above settings (and thats without whacking the core voltage and pushing it up to 690) his score would be in line but I suspect hes running the lowest driver settings possible.

[congo]

I realise there is only one GPU in existence, I just can't remember which one it is so ............

[NicksFXHouse]

I hate you. The best I can pull off is in the low 9,000s

But use Nicks advice to find out your optiomal.

Cheers
Cameron

9000 is easy with a 7900GTX 512



I ran that test and listened to some good tunes at the same time. No clocking the card, just let it putter along.

you must be using an Intel system

[flymo]

i willl max it out in 10mins and post the results again.
but now im gonna go do a FSX FPS test

[Ivan]

I just dont get it with all you guys freaking out on 3dmark. EVERY tweaker knows that only about half of the scene that 3dmark puts in the card is actually rendered, and the rest is just plain cheating.

[richardd43]

http://service.futuremark.com/compare?3dm05=2110203
3DMark05 = 12284

I just dont get it with all you guys freaking out on 3dmark. EVERY tweaker knows that only about half of the scene that 3dmark puts in the card is actually rendered, and the rest is just plain cheating.

As long as we all use the same software it allows us to check our tweeks for improvements and compare to other systems for possible improvements.

[NicksFXHouse]

I just dont get it with all you guys freaking out on 3dmark. EVERY tweaker knows that only about half of the scene that 3dmark puts in the card is actually rendered, and the rest is just plain cheating.

As long as we all use the same driver settings, the same standard 3D mark base test and are comparing the same manufacture of card, the test is valid

Although I do agree with the fact 3DMark is fixed to a certain degree, it does have some use.

[NicksFXHouse]

i willl max it out in 10mins and post the results again.
but now im gonna go do a FSX FPS test

Assuming a computer and Flight Sim is configured correctly, FS9 and FSX should both be LOCKED at 24-28 frames for high end cards... 22-24 for medium grade cards such as 6600's, x800-x850

Anyone who tells you different is totally clueless on how the rendering engine works for both products or is running FS9-FSX config settings that are not correct.