R423 will have 16 pipes(16x1) and will own nv40

[Carbon Unit]

Isnt the 9800 just a overclocked version of the 9700 (aaa)

[MUYA]

Essentially yes but apparently some stuff was fixed in the R350/R360 that wasn't working upto speed in the r3XX.

[-=DVS=-]

Isnt the 9800 just a overclocked version of the 9700 (aaa)

9800 got F-Buffer , 9700 does not (mag)

other then that its minor improvements with higher megahertz.

[StealthHawk]

9800 got F-Buffer , 9700 does not (mag)


But F-Buffer has never been exposed, so it's like a phantom feature, yes?

[Nv40]

But F-Buffer has never been exposed, so it's like a phantom feature, yes?


it seems so.. the last time i heard, it was not exposed even in ATi released drivers. its not anywhere in M$ Directx9 specs.. and this is weird .in ATI slides PS2.b the ones designed by M$ to support ATi incomming hardware they have no more "unlimited" shaders ,but the limit of 512. even less that NV3x hardware. by looking at ATi comments about it ,and others ,it seems that Fbuffer its too slow for games ,even in ATi hardware ,because it was more *software* based ,than hardware one.but also have read the it have some other limitations that make doubtfull its practical use ,in realtime games. so it seems that it was designed for specific situations like Offline renderings. havent heard of anyone using Fbuffer in opengl ,apart from ATI . if the feature have possibilies ,i am all for it ,but if it not practical for real use ,like gaming or true proffesional use ,then its best to move and let it die. we will know for sure with the R420 lauch ,if ATI will continue with that. but it seems very unlikely that any other IHV will support it ,specially if your hardware already support 1 THousand of intructions shaders. thats its more than enough for games at least until the next DIrectx10.
for offline rendering ,Professional market still have the QUadroFx boards that support more than 2k.

[-=DVS=-]

F-Buffer can only be used in OpenGl i think Nutty said or some programer.

[retsam]

wow muya now this all makes sence .... i never really belived in this whole 4,8,16 pipe stuff. it all makes the asic's sound like there running there clocks async ... just never made much sence to me now its starting to add up ....

[Nutty]

F-Buffer can only be used in OpenGl i think Nutty said or some programer.

Wouldn't know, Humus will tho.

[MUYA]

wow muya now this all makes sence .... i never really belived in this whole 4,8,16 pipe stuff. it all makes the asic's sound like there running there clocks async ... just never made much sence to me now its starting to add up ....


Dude I dunno 99% of the stuff ;) I just gathered what I could and tried to make head or tails of the whole thing.

Far more clever ppl around than me who can comment on the stuff with authority. :D

[Toaster]

ATI never released an extension to expose the F-Buffer in ogl. Guess they can't be bothered with it really since I doubt it will ever be used by any developer. The R420 will probably no longer have this feature.

[bkswaney]

It is no way near comparable to a r360. Just get this pipeline thing out of your head folks. As the r360 or r3XX doesn't have 8 pipelines. It has one, and so does the nv3X series. In fact Go search for tech info at b3d forums. All modern GPUs that DX8 compatible (i am quoting DaveB here) are GPUs that process quads in one pipeline. Therefore just saying or quoting a certain GPU acts like it has so much pipes and so much TMUs are not going to facilitate a apples to apples comparision. Just because you are hearing these inkwell stories about certain GPUs having certain configuration of pipes and basing your info on that is not tthe way to discuss relative performance to each other. This is what I have gathered from reading various posta nd correspondence with ppl

The R360/350/300 posses 1 pipeline which processes 2 quads per pass but thats all in theory when certains OPs are required it processes less. The NV3X is one pipeline too but the amount of quads it process is also governed by what OPs are required to be applied to pxels in the quad. The performance were lower for NV3X because of bubbles appearing in the pipeline due to penalties arising from certain conditions and hence many ppl deduced that was the reason for nv3X less than stellar PS2 performance compared to ATIs line.

Now NV4X is (rumoured) to be a heavily modified and essentially (rumoured to) capable of pushing 16 pixels per pass...that should equate to 4 quads. Thats all in theory and of course there will be inherent performance lost due to certain penalties when certain OPs or I think filtering techniques are called for but the feeling around the boards are that the new nv40 is more efficient and less bubbles are likely and that will equate to faster performance in PS.

R420 is a step from r3XX, it is also evolutionary (it seems though not much is known) than revolutionary just like the NV40. but how many quads it's able to prcoess per pass I dunno. Maybe 3 or 4. But the penalties are less for PS for r420 inherited from r3XX series than the nv3X and maybe NV40 but the amounts of quads being pushed the pipeline per pass is still limited by what OPs are required or to be processed. Therefore without any direct benches of theoreticals we cannot assess which will be faster. Lets wait till launch and when the previews come out with both both cards benched and then discuss those.

Just my 2c

Edit : of course it's more complicated than that but thats I think basic unerlying principles :p


OK... How does ATI get by claiming the R360 is a 8 pipe card? :confused:

[StealthHawk]

F-Buffer can only be used in OpenGl i think Nutty said or some programer.

It would hypothetically only be usable in OGL...if you could actually use it as it isn't exposed!

[MUYA]

OK... How does ATI get by claiming the R360 is a 8 pipe card? :confused:

Its not. It processes 2 quads at a time under peak/best conditions. So since it does process 8 pixels per pass under ideal circumstances, it's termed as 8X1 etc. ...

[Geforce4ti4200]

because the 9500 pro and up is really 8x1 pipes, thats why the 9700 pro at 8x1 owns the fx5950 at 4x2 cause dx9 games use all 8 pipes, dx8 games prefer 2 texture units so 8x1 is equal to 4x2

[MUYA]

GeForce4200, nobody is denying the performance numbers of r3XX over nv3X in particular PS 2.0 performance but your theory of why the performance is better is just not there.

[Sazar]

because the 9500 pro and up is really 8x1 pipes, thats why the 9700 pro at 8x1 owns the fx5950 at 4x2 cause dx9 games use all 8 pipes, dx8 games prefer 2 texture units so 8x1 is equal to 4x2

its not that simple m8...

[Geforce4ti4200]

arent pipelines the main strength of a card? 9500 pro>9600 pro despite 9600 pro being higher clocks

[Demirug]

arent pipelines the main strength of a card? 9500 pro>9600 pro despite 9600 pro being higher clocks

Yes, they are. But if you try to compare different chip and look only on the number of "pipes", TMUs per pipe and the clock you will not find the real answer. You can do this compare if you only look a chips of the same family. If you compare chips that are part of diffrent series you have to look much deeper.

nVidia and ATI use complete different internal designs for their pixelprocessors. ATIs desgin fit much better with DX9 because Microsoft likes it more than the nVidia solution. This is one part of the story the other part is that nVidia have to change to much of the NV3X design to fit it in the transistorlimit

[Evanescence]

This R420 ad on ATI Webpage is really FUNNY !

Reminds me for something like that:
http://www.ystart.net/upload/20040407/1081373628.jpg

[lokeshray]

err i dont see wats funny in that ad
maybe its just me,

by the way, in my opinion the new ATI logo sucks
i liked d previous logo....
lthe nvidia logo is nice

[MUYA]

Those Angela Jolie type lips are impressive .....

[Skynet]

Funny how lokeshray turned out to be spot-on in everything he originally stated.

[Evildeus]

Geforce4ti4200,
So the NV40 isn't a 16*1/32*0 part? :rolleyes:

[Sazar]

Funny how lokeshray turned out to be spot-on in everything he originally stated.

he is the reincarnation of nostredamus...

:eek: