Egg on ATI's face?

[MDolenc]

Is Floating Point Blending something possible in SM 2.0?

Blending doesn't have much to do with pixel shaders, so yes you can do floating point blending on PS 2.0 hardware (that supports FP blending).

 

[Ostsol]

Is Floating Point Blending something possible in SM 2.0?

I don't think that's part of the shader spec. It'd be more likely to be found in the specs for floating point offscreen buffers.

 

[Dave Baumann]

OK. AFAIK there are no qualatitive differences between the SM3.0 and SM2.0 paths in Far Cry. The SM3.0 path does use vertex instancung whuch will improve NV4x's vertex shader performance in comparison to not using it.

 

[Miksu]

Noticed this at Rage3D: http://www.pclab.pl/print9725.html

Couple quotes from the original thread:

Pawel Pilarczyk was one of the guests on this conference. ATI showed more then one sceenshot of HL2, they showed lots of benchmarks based on many popular games an advantage of R420 over NV40 was huge in the moments. He can`t post any details till early May, but from what I`ve heard in Tomb Raider AOD in High Res with FSAA R420 is almost 70% faster than NV40.

farcry running at 68 fps...I wonder what resolution/AA/AF settings were being used

 

[991060]

I'm sure R420 can beat NV40 when the latter is using 8xAA.

Does anybody know why nVIDIA didn't implement 6xAA? Is it hard to implement or protected by ATi's patent?

 

[digitalwanderer]

OK. AFAIK there are no qualatitive differences between the SM3.0 and SM2.0 paths in Far Cry. The SM3.0 path does use vertex instancung whuch will improve NV4x's vertex shader performance in comparison to not using it.

Deja-freaky-deeky-vu!

 

[Ardrid]

Noticed this at Rage3D: http://www.pclab.pl/print9725.html

Couple quotes from the original thread:

Pawel Pilarczyk was one of the guests on this conference. ATI showed more then one sceenshot of HL2, they showed lots of benchmarks based on many popular games an advantage of R420 over NV40 was huge in the moments. He can`t post any details till early May, but from what I`ve heard in Tomb Raider AOD in High Res with FSAA R420 is almost 70% faster than NV40.

farcry running at 68 fps...I wonder what resolution/AA/AF settings were being used

I'm skeptical of any information coming from sites that declare "allegiance" in their names, which is why I avoid nV News, Rage3D, and AMDZone.

 

[Rodéric]

Is Dave getting old to repeat himself ?

 

[Joe DeFuria]

I'm skeptical of any information coming from sites that declare "allegiance" in their names, which is why I avoid nV News, Rage3D, and AMDZone.

PCLab.pl has an allegience?

 

[Joe DeFuria]

Assuming that 70% faster in high res with AA is legit, I do have to wonder if that's, for example, comparing 8x modes.

But even if that's the case, it's still good news...I see this as evidence of one or more of the following:

1) ATI at least supports 8X MSAA mode
2) ATI has very efficient AA, or pipes in general
3) ATI has some significant bandwidth on board.

 

[Ardrid]

I'm skeptical of any information coming from sites that declare "allegiance" in their names, which is why I avoid nV News, Rage3D, and AMDZone.

PCLab.pl has an allegience?

Ahh...I completely forgot about that. The first thing that stuck out was that it came from the Rage3D forums.

 

[991060]

Joe, they probably compared R420 in 6x with NV40 in 8x, there's no 8xAA(MS, SS or whatever ) in R420.

 

[Joe DeFuria]

Joe, they probably compared R420 in 6x with NV40 in 8x, there's no 8xAA(MS, SS or whatever ) in R420.

Well, that wouldn't really be a fair comparison if that's what they did. I'm hoping it's not.

BTW...since when has it been known that ATI doesn't have 8X AA?

 

[AlphaWolf]

Joe, they probably compared R420 in 6x with NV40 in 8x, there's no 8xAA(MS, SS or whatever ) in R420.

You know this for a fact, or are you just assuming that?

<edit> and there are cases where the 9800XT could say its faster if it was 6xAA vs 8xAA, so I rather doubt that was the comparison.

 

[PaulS]

Interesting stuff 8)

 

[991060]

You know this for a fact, or are you just assuming that?

Let's say it's a fact waiting to be confirmed.

 

[Evildeus]

Is Floating Point Blending something possible in SM 2.0?

Blending doesn't have much to do with pixel shaders, so yes you can do floating point blending on PS 2.0 hardware (that supports FP blending).

Thanks for the clarification (and Ostol also )

 

[Joe DeFuria]

You know this for a fact, or are you just assuming that?

Let's say it's a fact waiting to be confirmed.

OK...but if it turns out to not be the case, you get 50 lashes with a wet noodle!

 

[991060]

OK...but if it turns out to not be the case, you get 50 lashes with a wet noodle!

Yes, I know anything could change before we see the silicon, but this thing is just not gonna change. For me, it's not more likely for R420 to have 8xAA than NV40 to have 6xMS.

 

[Evildeus]

Joe, they probably compared R420 in 6x with NV40 in 8x, there's no 8xAA(MS, SS or whatever ) in R420.

I doubt that, the 6800U @8* is already 30-50% slower than a 9800XT @6* (meaning 50-200% faster)
http://www.xbitlabs.com/articles/video/display/nv40_15.html