GF FX vs ATI

[Razorscott2YK]

As it seems, most modern games these days seem to be going more towards Direct 3D these days so your best bet at this time would be the ATI cards. I used to be a huge Nvidia Fan with my outdated Geforce 4 Ti4200 until i saw just how smoothly and how much better an ATI card would run Halflife2 and other DX9 based games. So a few weeks later the TI4200 was laying in my spare parts box and a Brand new 9600XT "Xtreme" edition card was gracing my computer with it's almighty performance. No offence Nvidia, but if your cards are "technically" superior and STILL outmatched by a "technically" slower card, then you've got some serious ass kicking to do in your tech department.

Oh and you get a free version of HL2 when it comes out if you get an XT, though i think you have to download it which will suck if you have only a 56K modem.

But eh, it's your choice.

Cheers.

-Razor2yk

 

[jonbob]

I would like to know what you are refering to when you say "they have been known to cheat too - just not as much". If its the quak cheat, it was so long ago nobody hardly remembers it.

I forget the details, but something to do with the ATi cards using lower quality textures in place of the correct ones, and something to do with Aquamark3. Not that it matters - ATi cards still offer far better performance than equivalent-priced nVidia cards. (At least in the current crop of games - but who knows what tomorrow will bring with the likes of the new Doom engine and Serious Sam 2?)

 

[dawdler]

I forget the details, but something to do with the ATi cards using lower quality textures in place of the correct ones, and something to do with Aquamark3. Not that it matters - ATi cards still offer far better performance than equivalent-priced nVidia cards. (At least in the current crop of games - but who knows what tomorrow will bring with the likes of the new Doom engine and Serious Sam 2?)

The low quality texture problem in UT2k3 is a program BUG, not a cheat. Its easy to fix. The Aquamark3 gamma problem is global, and confirmed by both independant tech gurus and ATI driver devs to not affect speed at all, its simply the way ATI renders it (and that is from versions far before the R300 generation), slightly darker than Nvidia.

Questions?

 

[jonbob]

Oh well, looks like I might have been mistaken. But I still say that I find it hard to believe that in an industry as competitive as this one a company like ATi has done nothing wrong at all. Incidentally, what caused the Halo problem? I have to admit it was starting to irritate me.

 

[Arno]

The Aquamark3 gamma problem is global, and confirmed by both independant tech gurus and ATI driver devs to not affect speed at all, its simply the way ATI renders it (and that is from versions far before the R300 generation), slightly darker than Nvidia.

I thought the Aquamark3 differences were related to an alphablend bug in the R300 hardware, as described in this thread ?
Quote from that thread (supposedly a direct response from ATI):

It took much longer than I anticipated, but we reached to a conclusion on the AM3 "explosion" issue. The initial suspicions about our alpha blenders turned out to be true. To prove this we had to run the test scene on a future HW emulator - which took all this effort.

To summarize at a high level - we have single-bit differences when alpha blending compared to ref-rast image. When accumulating several layers of alpha blend objects, some pixels may end up looking darker on R300 family than refrast. We are rendering all the layers and the deviation we have in rendering each pixel is within an acceptable tolerance level from refrast - as evidenced by the fact that we pass all the WHQL DCT tests that compare images rendered on R300 v/s refrast.

So basically the Aquamark3 issue is related to alphablending and only occurs on R300 hardware.

 

[dawdler]

Aaah, I need to update myself
I went with what they said earlier, this is later info... You are of course correct Arno, sorry. (I dont see why it isnt global though, AQ3 isnt the only game that does alpha blending, is it? And the entire scene was darker too). With it predating the R300 and the issue of cheating, I was refering to the drivers (the devs noted earlier that all drivers rendered it the same). Its not a cheat as in deliberate driver modification (in any version), its just the way it renders, albeit a bug hehe.

 

[Arno]

No problem, I'm glad I could share some useful info.
The alphablend bug is in fact a global bug. The first post in that Beyond3D thread I linked to shows a simple example program that identifies the bug. It only becomes noticable when a lot of transparent layers are rendered on top of each other and that's the reason why you don't see it happen a lot. AnandTech noticed an issue with the lightsaber in Jedi Academy, which could be related to this bug.

And, yeah, it's not a deliberate cheat. It's just a small error in the R300 hardware. It will be fixed in the next generation of ATI graphics cards.

 

[jonbob]

Speaking of which, when is the new line expected?

 

[63]

I have a Radeon 9600 Pro AiW, and it works great! I paid $299 CAN