Ati accuses nvidia of cheat

[harl]

http://www.theinquirer.net/?article=15502

They say Ps3.0 in far cry are equal to ps2.0

That's begining :ORDER:

[oqvist]

Yeah it explains why nVidia didn´t want to compare it against a video card running in PS 2.0 mode. Noone would ever see the difference :D

Another proof of how much those PR launches are worth :D And people complain about people downplaying PS 3.0??

[shinrai]

I love it!

It really didn't take long for nvidia to stick there foot in it.

[lIqUID]

go to a car lot and ask the Chevy dealer if the new Ford trucks are better than theirs and see what answer you get.

[mikez]

awww, boooooohoooooo. let's all join in ATI's circle and think happy loving thoughts so that big bad NV demons will not intrude the circle of peace.

go back to your bongs hippies...

[DSC]

Funny, future red company's hardware will support SM3.0, but of course atm lets use FUD against the competitor to sell more of our same, rehashed hardware since 2002.

:lame:

[The Hip]

So that means FX owners will get a huge IQ improvement once they upgrade to GF6 series. Well at least people that own the company red's rehashed hardware don't need to upgrade to get similar IQ.

Yup - that is really lame.

And look at the following snippet:

Quote:
"A CryTek representative responded on this matter with this answer: "Was Nvidia showing SM3.0 vs. SM2.0 or SM1.1?" He replied to his own question by saying that Nvidia was showing 3.0/2.0 vs. 1.1."

So WHO downplay SM3.0 here... I believe it the representative of CRYTEK that put SM3.0/SM2.0 in the same category - but then again what do the developer know.

[retsam]

So WHO downplay SM3.0 here... I believe it the representative of CRYTEK that put SM3.0/SM2.0 in the same category - but then again what do the developer know. well i think you missed the significants of his statment why code for ps 3.0 if 2.0 does the same thing .... there is a reason that there coding to ps. 3.0 also .... one is that it takes less code to do the same thing and two from what i gathered from reading b3d its more effience then ps 2.0 ....


just my thoughts

retsam

[mikez]

there is a reason that there coding to ps. 3.0 also .... one is that it takes less code to do the same thing and two from what i gathered from reading b3d its more effience then ps 2.0 ....

is that so? can anyone confirm it? if the difference is code efficiency then NV's PS3 quite obviously wouldn't render anything different in PS2 mode but the improved efficiency would mean faster rendering wouldn't it? a speed difference comparison between PS2 and PS3 would be good if that's the case.

[retsam]

anyone have the link to the nv40 launch parth they did say something in there about some mode with the space ships and stuff ...... anyone remember that

[oqvist]

is that so? can anyone confirm it? if the difference is code efficiency then NV's PS3 quite obviously wouldn't render anything different in PS2 mode but the improved efficiency would mean faster rendering wouldn't it? a speed difference comparison between PS2 and PS3 would be good if that's the case.

It would if not Shader model 3 required 32 bit precision in the specs ;)

I do think ATI:s 24-bit PS 2.0 extended is considerable faster than the 6800 ULTRAs 32-bit PS 3.0... I don´t think the higher efficiency of running longer shaders make up for the higher precision...

But perhaps it will allow nVidia to take the IQ crown if ATI don´t improve over their R3XX cards??

[retsam]

http://www.gamers-depot.com/interviews/dx9b/001.htm


it sounds that all the developers are saying the same thing that im saying p.s 3.0 is more efficent then ps 2.0 the same things can be done but its just all around easier... its a funny read but the devoplers that seem to explian ps 3.0 in detail seem to have a positive outlook on it.... but when a developer doesnt go into any detail about ps 3.0 they arent to keen on it



retsam

[oqvist]

Programmers always love when people make things easier on them don´t they. They sure hated cg that I can tell you. The FX was apparently a nightmare to programme for and thus I guess it´s just about only DOOM 3 that will be able to use it´s full potential since it was as much John Carmack that produced the NV30 core as nVidia.

[retsam]

It would if not Shader model 3 required 32 bit precision in the specs

I do think ATI:s 24-bit PS 2.0 extended is considerable faster than the 6800 ULTRAs 32-bit PS 3.0... I don´t think the higher efficiency of running longer shaders make up for the higher precision...

But perhaps it will allow nVidia to take the IQ crown if ATI don´t improve over their R3XX cards??

your right on that note it is a traid, but it sounds like its an even trade off at worst .. 32 floats take more but the code effiancy makes up for it.....and also the precision can be applied at the devoplers own will ... fp 32 and p.s 3.0 is all just another choice for the dev's .....

[eesa]

uhggg.... I'm not going to touch on speed as that has yet to be seen in the real world, but for the last time there's effectively nothing in 3.0 that you can't do with 2.0. Even go and reasons they have for 3.0 over 2.0. I'm not impressed. Sounds like they're trying too hard.

[retsam]

eesa.... have you even read that artical the one i linked up there .... didnt you notice the devs that seemed to explain in great details of ps 3.0 seem to love it ...... i just think the whole ps2 can do the same as ps 3.0 is a myth ....


mark from nova logic.....
Mark: Absolutely! The ability to perform dynamic flow control opens the door for many new things, such as true per-pixel ray tracing, and support for arbitrary numbers of light sources at the per-pixel level. Also, shaders can be greatly optimized by careful usage of conditionals and predicates (partially available in 2.x but not at all 2.0



Andrey Khonich cyrtek makers of farcry

In VS3.0 Shader model actually is possible to support general displacement mapping (with smart Shader design when vertex Shader has to do something during waiting for texture access).


and this

- Handle several light sources in single pixel shaders by using dynamic loops in PS3.0.
- Decrease number of passes for 2-sided lighting using additional face register in PS3.0.
- Use geometry instancing to decrease number of draw-calls (remove CPU limitations as much as possible).
- Unrestricted dependent texture read capabilities to produce more advanced post-processing effects and other in-game complex particles/surfaces effects (like water).
- Full swizzle support in PS3.0 to make better instructions co-issue and as result speedup performance.
- Increase quality of lighting calculations using 32 bit precision in pixel shaders on NV40.
- Take advantage of using 10 texture interpolators in PS3.0 Shader model to reduce number of passes in some cases.
- Easily do multiple pass high-dynamic-range rendering on FP16 targets.
- Speed up post-processing a lot by using MRT capabilities and mip-mapping support on non-power-of-two textures.

and then there is tim sweeny i love his comment

Tim: Only a marketing guy would consider 24-bit floating point to be "full precision"!

32-bit floating point has been an industry standard for 18 years, since the original IEEE 754 specification was ratified. Besides the obvious need for 32-bit precision in any sort of high-range calculation (such as position vector math, world transformations, etc - all things one is likely to do frequently per-pixel), we feel there's a huge advantage to using industry standard arithmetic in its predictability and reproducibility across different GPUs. The 24-bit precision mode allowed by the DirectX9 spec was backed by a reasonable cost/performance/precision judgment back in the early days of DirectX in 2002-2003. But nowadays there isn't a good reason for hardware to support less than full 32-bit floating point precision.


hmmm seems these guys proved my point


retsam

[jbirney]

retsam,

both of those developers are being paid by NV since both are members of TWIMTBP program. Anytime a paid person offers you their opinion you have to wonder how much is the truth...how much is...well paid for :)

Notice how the developer that are not on the program did not think PS3.0 was "all that and a bag of chips"?

[The Hip]

@retsam

Didn't you "miss" some developers statements that was not in the TWIMTBP.

Else your browser have mighty fine filtering.

SM3.0 is more efficient than 2.0, but IQ wise everything that is possible with 3.0 is also possible with 2.0.

Tim Sweeny .... (no comment).

[vampireuk]

ATI = :argh: :cry:

[Arrghman]

TWIMTBP aside, that's absolutely correct about 32-bit. Software rendering packages that are full percision rendering engines all have 128-bit pixels (32 for each of the 4 channels).

[jimmyjames123]

ATI claims:

ATI also claims that its hardware will run faster then Nvidia's anyway and he added: "It's pretty much impossible to make a sm3.0 game look noticeably different to a sm2.0 game, which is why Nvidia was comparing the 2.0/3.0 path with a 1.1 path."

The reason NVIDIA would compare PS 3.0/2.0 to PS 1.1 is because they were comparing the NV40 to the current generation NVIDIA FX cards, which make use of the PS 1.1 effects.

It's much ado about nothing, really.

[McDusty]

"The ability to perform dynamic flow control opens the door for many new things, such as true per-pixel ray tracing, and support for arbitrary numbers of light sources at the per-pixel level."

- ive written a ray tracer on my 9800pro using a multipass system. Yes this could be done in a single pass (or less at least) with PS 3.0, but just because something requires multiple passes it doesn't mean its slower.

"In VS3.0 Shader model actually is possible to support general displacement mapping"

- hes talking about VS 3.0, not PS 3.0. VS 3.0 does have something to offer.

" Handle several light sources in single pixel shaders by using dynamic loops in PS3.0."

- Multipass

etc, etc, etc

Tim Sweeny's work doesn't require FP24 let alone FP32 and won't do until 2006. The guys at epic are in Nvidas pocket and I don't think anything they have to say can be considered objective.

[The One]

retsam,

both of those developers are being paid by NV since both are members of TWIMTBP program. Anytime a paid person offers you their opinion you have to wonder how much is the truth...how much is...well paid for :)

Notice how the developer that are not on the program did not think PS3.0 was "all that and a bag of chips"?

lol yea cuz the other dev's are in the GITG program from ati

i find it funny that some people think so highly of ati, they are as bad as nvidia they both lie to try to sell their product

[eesa]

eesa.... have you even read that artical the one i linked up there .... didnt you notice the devs that seemed to explain in great details of ps 3.0 seem to love it ...... i just think the whole ps2 can do the same as ps 3.0 is a myth ....



actually yes, I read them. That's why I was commenting on what they said. If you look at the things they're pointing out, they're trying really hard to make it seem worthwhile.

[eesa]

anyway, you know what? I just realised something. I really don't care about this :) I probably will care for the generation after, but right now you could support PS5.8797 for all that gets you. Even if it's implemented in a few games now, I don't see that huge of an impact. Assuming that PS3.0 is worth its while, all good features start off as a checkbox feature and eventually have enough hardware backing them up that it becomes really useful, and at that time the games will be out that take advantage of it. I don't believe that time is now.