R420 Hints from people under NDA!

[Diplo]

Don't underestimate peoples' lust for the latest features. Men (and let's face it, the vast majority of people who buy the latest graphics cards are men) have always had a weakness for hi-tech gadgets, no matter how impractical they actually are.

I read quite a few different gaming forums, and every single one of them has featured numerous posts about the shakey-cam footage of the Unreal Engine 3 footage. People who had no idea what a shader was are now familiar with the term pixel-shader 3 and equate it with the jaw-dropping graphics they saw. More importantly for Nvidia they equate it directly with the 6800 and 'know' that the next ATI card won't have it.

Now, must of us realise that graphics in the Unreal video will most likely never be realised within the working lifetime of the NV40, but will that stop people lusting after it's 'killer' feature? Nope!

 

[anaqer]

People who had no idea what a shader was are now familiar with the term pixel-shader 3 and equate it with the jaw-dropping graphics they saw. More importantly for Nvidia they equate it directly with the 6800 and 'know' that the next ATI card won't have it.

Exactly.
Even if PS3.0 ends up being a checkbox feature or a developer's toy it will still sure as hell contribute a LOT to NV40 sells.

 

[Kombatant]

If that's true and ppl lust for new features, I am sure ATI could pull something off in that regard

 

[Mulciber]

there's nothing i hate more than winkers

 

[jolle]

If that's true and ppl lust for new features, I am sure ATI could pull something off in that regard

isnt R420 said to use some extended version of PS2.0?
hasnt been alot of talk about that, mostly SM3.0 stuff latly..
i want info!

 

[martrox]

First of all.....No one here really knows and has said exactlly just what the R420 will or won't have. It "might" have 3.0, it "might' have 2.0++, it "might" have 2.0 only and it "might" have a lot more usable speed.... big term here is "might", because you don't know!!

One thing we do know is the R420 will have better FSAA...... because the r3xx does NOW. And it will have as good AF.... because the NV40 has dropped to a version of adaptive AF......

All of you here - and you know who you are - that, at this time, keep bringing up the "features" that the NV40 has that the R420 doesn't are - how can I say this without personal attacks or insulting remarks - (oh, well, sorry JR, I can't) are either so biased towards nVidia or are just plain morons.... because you just don't know, period.

Nothing I have written is to in any way demean the NV40's abilities - hell, I very well "may" end up owning one or two. It has some great features.... and is missing at least one, better FSAA, which for some is a "big" thing. And it sucks some serious power. Again, a big thing for many. These are things we do know.

And, to go back to the few here I "may" have insulted..... arn't some of you the same people that said how great and how far ahead the NV3x was/is?

 

[Kombatant]

First of all.....No one here really knows and has said exactlly just what the R420 will or won't have. It "might" have 3.0, it "might' have 2.0++, it "might" have 2.0 only and it "might" have a lot more usable speed.... big term here is "might", because you don't know!!

Actually that is not entirely accurate. Some people already know that information; I for myself have dropped some hints the past few days. It would be (I believe) rather immature on my account (I can only talk for myself, as always) to start talking out of my *insert word that'll probably get censored here* about something I have no solid information on.

 

[Diplo]

First of all.....No one here really knows and has said exactlly just what the R420 will or won't have.

We know that, but there are a lot of people out there on the gaming forums who think Nvidia are going to be the only ones with PS3.0 They may be right, they may be wrong, but the fact they believe it is a big victory for Nvidia. Remember, facts have little to do with marketing, and Nvidia consistanly beat ATI in one respect - marketing.

 

[Sxotty]

Just need another group of vague statements from people supposedly in the know. Where is HB when you need him anyway.

Oh you just wait for the r420 you'll thank me who says that is all they will have bla bla blah To bad people just leave you hanging and in the end if they are completely wrong they will just point to their vague statements and go "look I never said it would be better just that you should wait and look you saved $$ for waiting", or "look it does have pointless feature Z" now we can all jump for joy at the inanity of it, or perhaps the famous "I was right"

Then these people criticize the Inq in this fashion "With so many statements they have one for every outcome and thus cannot be wrong" Well these same people are afraid to make a definitive statement and be wrong themselves so meh...

 

[Kombatant]

Just need another group of vague statements from people supposedly in the know. Where is HB when you need him anyway.

Oh you just wait for the r420 you'll thank me who says that is all they will have bla bla blah To bad people just leave you hanging and in the end if they are completely wrong they will just point to their vague statements and go "look I never said it would be better just that you should wait and look you saved $$ for waiting", or "look it does have pointless feature Z" now we can all jump for joy at the inanity of it, or perhaps the famous "I was right"

Then these people criticize the Inq in this fashion "With so many statements they have one for every outcome and thus cannot be wrong" Well these same people are afraid to make a definitive statement and be wrong themselves so meh...

If you believe someone is not credible, then you don't have to give validity to his/her comments, that's a given I believe. Those who are in the know have NDAs signed so they cannot reveal anything; just drop hints. I bet these people would love to reveal what they know, but they can't because of the legal implications. So could you ever expect them to make a definitive statement?

 

[Sxotty]

Yes I expect them to make a definitive statement such as this


That is what an NDA is for. If they cannot say something specific such as "I don't beleive it has two molex connectors" Then IMO it is of little to no value in any way except to make them feel important. Everyone likes tidbits of info, but they don't get it from these posts that say nothing concrete anyway. At the worst case if you feel you must say something to save the whales (err masses) from buying the latest and greatest from whomever b/c you think it will be obsolete, then simply say "The XX is coming out in 3 weeks and it is a good idea to wait until all the cards (heh punny) are on the table.

edit: BTW have you seen the movie "out of time" your post reminded me of the scene in the bar "Well chris maybe this guy would like nothing better than to tell you I'm boinkin your wife, but she won't let him" Except it was like Well sxott those under NDA would like nothing better than to tell you what they know... sorry just found it funny.

 

[Kombatant]

You mean make a blank statement? :

 

[jolle]

maybe he meant

There is no hidden message in the statment

 

[Doomtrooper]

Hehhe

 

[jolle]

Hehhe

hehe, well there wasnt, i checked..

 

[Doomtrooper]

First of all the above argument is what 3dfx loved to use "nevermind the features, look at our speed." Look where it got them. Regardless of what certain people in this forum like to think, its a proven fact that features sell products, especially graphics cards.

If your analogy was correct the Original Radeon should have outsold the GF2 and the 8500 should have outsold the GF4. Who are you trying to kid here, people upgrade their graphic for speed reasons 99% of the time. When they make the decision to upgrade they turn to magazines and web reviews that spend 99% of the press on speed graphs.

 

[KimB]

Those who are in the know have NDAs signed so they cannot reveal anything; just drop hints.

The thing is, there are those at ATI who are not bound by such agreements. If ATI was to support PS 3.0 on the R420, you would expect that we would have heard from a higher-up from ATI by now that it would.

Instead, we have ATI at GDC talking about this "PS 2.0b" profile, a profile which no current hardware supports, and no denial of lack of PS 3.0 support.

It seems apparent to me that PS 2.0b sums up the R420's capabilities pretty well. The only question remaining is whether or not it will support longer programs in OpenGL (the PS 2.0b supports the maximum number of instructions that PS 2.x allows).

 

[jimmyjames123]

I think you may be right about that, Chalnoth. Logically, it would not make sense for ATI to keep so quiet about PS 3.0 if their chip actually did fully support all PS 3.0 functions, especially now that NVDA is really pushing PS 3.0 capability. But then again, obviously no one can underestimate ATI. We will see soon enough.

 

[Kombatant]

Those who are in the know have NDAs signed so they cannot reveal anything; just drop hints.

The thing is, there are those at ATI who are not bound by such agreements. If ATI was to support PS 3.0 on the R420, you would expect that we would have heard from a higher-up from ATI by now that it would.

Instead, we have ATI at GDC talking about this "PS 2.0b" profile, a profile which no current hardware supports, and no denial of lack of PS 3.0 support.

It seems apparent to me that PS 2.0b sums up the R420's capabilities pretty well. The only question remaining is whether or not it will support longer programs in OpenGL (the PS 2.0b supports the maximum number of instructions that PS 2.x allows).

I see what you're saying man. But bear in mind that ATI hasn't confirmed or denied anything, including the thing that has been going around about R420's lack of PS3 support. So it wouldn't be wise to base your opinion solely on that fact methinks.

Of course by that I am not saying that you are wrong in your assumption (or right for that matter). Just pointing out something that strikes me as odd in the logic you used.

 

[jolle]

IF it turns out that they dont support it, the next question will be what role
will SM3.0 play in the timeframe between now, and when they do give out
a SM3.0 part..

I see it about the same as with 64bit computing..
Its not vital, AMD took the step cause it would have been a natural evolution
that would occur sooner or later, atm there isnt much softwaresupport, and
intel will use 64bit extentions when the software is out for it..
But if AMD wouldnt have taken the first step, the software wouldnt be in
the works as we speak...

in the near future the role of SM3.0 might not be a vital one, just a neato extra
feature, even tho it has potential to be more..
and will be once it gets used in games specificly created for it..

Dunno tho.. but alot of takes on it sounds like either overestimating or
underestimating it, or rather the role it will play in the near future..
features usually dont become vital until the hardware support is somewhat
mainstream on the market..