What shall I do with my Geforce FX?

P

pocketmoon_

Newcomer
I've just recieved some news from Nvidia - they are now shipping the cgshader.com contest prizes. So... that's 4 GeforceFX's going to Brazil, Brazil, Russia and the UK this week. I hope to get mine before the weekend. :)

So, the question is...

What shall I do with my Geforce FX ?

Serious (benchmarks) and not so serious (scare cats) will be concidered

[Edit]
I have a P4 1.8 and an original GF3 but I'm sure I can throw some synthetic benchmarks at it. One think I want to test is half vs float shader performance.

[Edit2] I have to say I'm exceedingly happy to be getting one of these.
 
And the FXFlow cracks continue. :)

Seriously though, congrats on winning the FX!

What you'll find here, is that most people are less interested in actual performance numbers (because we'll get them from everywhere else), but in digging deeper to try and find out how particular features have been implemented.

If you do some searching on this forum (and read some of the recent reviews on this site), you'll find links to some programs that assist in analyzing anisotropic quality and FSAA sample position.

Aside from that, any benchmarks that you can run on "DX9 Shader" tests that others can also run on the 9700 from comparison....that would be great.
 
First of all congrats on winning one of the prizes pocketmoon! If you conduct some synthetic benchmarks be sure to inform us of any unusual or interesting findings. Some IQ impressions (AA/AF) would be most appreciated too...
 
I'd like to see benchmarks using 6xS and 8xS FSAA.
If you could use aTuner from 3dcenter to enable the pure 4x SSAA modes that would be great too.
Of course combined with 8x Aniso.

If you have the doom alpha some comparisons would be nice. ;)
 
Speaking of Doom...I thought it was rather odd that no mention was made about Doom III performance on the FX. I seem to recall a certain marketing PDF that claimed how much faster the FX was than the 9700...of course, no mention whatsoever on what the actual demo/benchmark was, the settings involved, etc.
 
OK, the GeForce FX 'funnies' now have their own thread here. Post the images there if you want. Lets keep this topic on track please.

(Oh, but my suggestion was to post it to me for a week or two of testing! ;))
 
We're not talking about the leaked alpha.

Back at Comdex, NVIDIA released some "benchmark scores" for the NV30 and the Radeon 9700. They had no problem distributing that information to web-sites to publish.

Now that the NV30 is in "reviewable" state, it's only fair to at least disclose how the test was performed, and allow web-sites to verify the results.
 
doom III - well I'll see what enabling the NV30 bit in the config file does ;)


I have been told that the NV30 path in the leaked alpha is way way out of date!
 
pocketmoon_ said:
doom III - well I'll see what enabling the NV30 bit in the config file does ;)


I have been told that the NV30 path in the leaked alpha is way way out of date!
Click to expand...

Well, John Carmack has stated before that they have a NV30 backend ans only an R200 backend. I wonder what they are doing for the NV30 and how that fact will influence the performance spin betweeen R300 and NV30.

Maybe nVidia finds some comfort in knowing that Carmack will be nice guy to them. ;)
 
Ante P said:
Oh another thing, please run Vulpine GL Mark.
Click to expand...

It doesn't seem to tell you anything conclusive in my experience. Now I am not saying this because all the NV cards seem to shine in the benchmark and all non-NV cards don't. I am saying that we don't know 'why' it is the case when other benchmarks dont show the same performance differential no matter what generation of NV or non-NV card.
 
misae said:
Ante P said:
Oh another thing, please run Vulpine GL Mark.
Click to expand...

It doesn't seem to tell you anything conclusive in my experience. Now I am not saying this because all the NV cards seem to shine in the benchmark and all non-NV cards don't. I am saying that we don't know 'why' it is the case when other benchmarks dont show the same performance differential no matter what generation of NV or non-NV card.
Click to expand...

I mainly want him to first test without the Vertex Shader and Pixel Shader features and then with them on.
 
You must log in or register to reply here.
Back
Top