3DMark 2003

[Scouse]

Have a look at this Xavier m8:

Futurmark's Response to Criticism



Kan m8 - if your games are running fine then you don't feel the need to upgrade. If your games are running shitty you tend to use benchmarking programs to find out what part of your system is running like the proverbial "dog".

You don't run a benchmark to decide if you need to upgrade - you decide you need to upgrade then you run your benchmark to find out what you need to upgrade.....


.....or am I talking shit?

 

[kan]

Originally posted by Scouse

.....or am I talking shit?

dunno i lost interest about a page ago

i see what your saying tho

 

[bodhi]

Xavier, face it, GeForce FX is a piece of shit. Over-delayed, over-hyped, over-priced, shitly designed, under-performing, and what's more the Ultra has been canned.




Sucks to be you man.

 

[Durzel]

GeforceFX does indeed appear, on face value, to be a bag of wank.

Not sure how Nvidia managed to create a card which actually comes out slower than a Radeon 9700 Pro in virtually all tests.

ATi really are leading the pack at the moment. Wonder how long that will remain the case?

 

[Dimebag]

Parp

 

[Durzel]

<Dimebag> I've got a new bf now, his name is steve

 

[Dimebag]

Where did that come from huh?

Bit unprovoked!

 

[Xavier]

Originally posted by Durzel
3DMark2003 is a DirectX 9 benchmark.

Geforce Ti4600 uses pixel shader version 1.1, vertex shader version 1.3, and doesn't support DirectX 9 (no optimisations for it)

Radeon 9700 uses pixel shaders version 2.0, vertex shader version 2.0, and is optimised for DirectX 9.

Gaming benchmarks are heavily GPU-loaded (since the graphics card does about 75% of the work).

So are the results that surprising?

NVIDIA use PS1.3 for DirectX8
SIS use PS1.3 for DirectX8
Matrox use PS1.3 for DirectX8
ATI use PS1.4 for DirectX8
Both use 2.0 for DX9

noones talking about testing DX8 cards on this I think - the problems have arisen because the codepaths ignore the single-pass solution for NVIDIA/SIS/Matrox hardware (1.3) and only use 1.4 falling back to 1.1, so unless you're ATI you're rending 4x as much... something which no DX8/DX9 games will do.

Don't forget, 1/4 of the shaders in 1/4 of the game tests in 3DMark'03 are actually DX9 (so about 1/16th of the code tested) - so much for 'The Gamers DX9 benchmark'

 

[Xavier]

Originally posted by Scouse
Aye Xavier - but don't forget that until last December Nvidia were one of the design partners of 3Dmark03.

It kinda beggars belief that Nvidia would go along with the design team for so long without pointing this out don'tya think.

It's also suspicious that they pull out a matter of weeks before launch.

And Nvidia cards from now on will support the DX9 stuff - so I think it's really more about 3dmark getting great scores on the current ATI lineup.....

Heh, it wasn't just NVIDIA who objected to the PS1.3 absence in the benchmark code, Matrox and SIS also use the same standard, it's only ATI who can do any of the DirectX8 shaders in a single pass, which are responsible for over 80% of the shader code in the entire benchmark...

 

[.cage]

WHO GIVES A FUCKING TOSS

I IMPLORE YOU TO DO SOMETHING PRODUCTIVE

 

[Testin da Cable]

xav is doing something productive cagey. different people do different things. sadly, we can't all be students

 

[.cage]

i dont mean xav in particular, but the whole
I GET 100 MORE 3D MARKS =) HUHUHUHUUH thing is very, very stupid :|

 

[Testin da Cable]

well, I'll give you that as I agree. I'm a full sub to the 'if it plays ok then there's nothing wrong with it' school atm.