More ATI Missing Vertex Texture Fetch Info

Status
Not open for further replies.
pxc said:
Not necessarily a performance hit. Most likely it will just not show the effects linked to VTF. No one is going to code the workaound into a game if the x1800 refresh (R580?) includes VTF.
Click to expand...

Again, you're guessing and you have no information that points one way or another. They could not use VTF (as everyone has not except for one game) which is the most likely scenario or they could develop codepaths for each card whether because they just do or ATI throws cash at them like nvidia does to developers.

Unless you work at ATI and major game developers and somehow know what the X1800 referesh is going to have and what developers are going to do your statement is complete shot in the dark.
 
tornadotsunamilife said:
I don't see why not, from what I see unified architectures doesn't need VTF anyway and that's the direction that ati is moving in (if not nvidia also)
Click to expand...
SM3.0 isn't unified. :p
 
OldBoy said:
Again, you're guessing and you have no information that points one way or another.
Click to expand...
Sorry that I can't paint a picture of unicorns and rainbows. :rolleyes:

Vendor specific hacks are usually ignored for standards. Look at history.
 
pxc said:
Sorry that I can't paint a picture of unicorns and rainbows. :rolleyes:

Vendor specific hacks are usually ignored for standards. Look at history.
Click to expand...


Sorry that once again you're just making things up. VTF isn't supported by anyone except for one game and even that game has issues with it:

http://forums.ubi.com/groupee/forums/a/tpc/f/49310655/m/2531042843/p/1

Stop doing your nostradamus routine as if you know any more about the future than anyone here. You simply don't know and your assumptions are a waste of everyone's time.

If you want to look at history, then notice that features that have little to no real world applicability rarely if ever get used.

Finally, save your smart ass comments for yourself. If you're not going to act like a civil adult go elsewhere.
 
PRIME1 said:
It's supported by the SM3.0 spec. That's all that matters for this thread.
Click to expand...

Do me a favor and stay out of a tangential discussion if you're not going to address that discussion.

Your point is worthless in the discussion subset me and pxc are having.
 
PRIME1 said:
It's supported by the SM3.0 spec. That's all that matters for this thread.
Click to expand...

the method ATI uses seems to be better =p maybe you should read? also according to what the articles and forum posts are saying ATI is within SM3.0 Specs even without HW support for VTF
 
It doesn't really matter now and most likely in the future as well, any performance hit it takes is going to, mostly likely, be negligible.

Now what does matter is ATI's marketing as "SM3 done right" when they're missing a part of SM3s feature set and how funny it is. Much like one of NV's "Way it's meant to be played" games running faster and looking better on an ATI card.

LOL
 
CrimandEvil said:
Now what does matter is ATI's marketing as "SM3 done right" when they're missing a part of SM3s feature set and how funny it is. Much like one of NV's "Way it's meant to be played" games running faster and looking better on an ATI card.
Click to expand...

Like far cry?
 
My question still hasn't been answered. Lets pretend we are at dx10.
Also lets pretend that we are in the future and both nVida and ATi have gone to a completely unified archiecture.
In a unified architecture, task specific units in GPUs will be greatly reduced, or not there at all. If software tells the unified architecture to DO specific dx10 instruction in hardware, do they meet dx10 API reuirements?
If they do, then it stands to reason that if software tells parts of the r520 to do what is missing as a specific unit in the r520 hardware on another part of it's hardware, it is then compliant? And if not, then no unified architecture will meet any API requirements. Or am I totally off?
 
PRIME1 said:
It's supported by the SM3.0 spec. That's all that matters for this thread.
Click to expand...
Pacific Fighters also uses it. There might be other games, but that's the oldest game that uses that part of VS3.0.
 
Digital Viper-X- said:
the method ATI uses seems to be better =p maybe you should read? also according to what the articles and forum posts are saying ATI is within SM3.0 Specs even without HW support for VTF
Click to expand...
I have seen no proof that their method will even work. Since it's out of spec developers will have to program a seperate path for it, or ATI will have to translate the instructions with in the driver. That does not sound better to me.
 
If ATi feel they have to include this workaround feature then I'm sure there will be a driver released with it included
 
OldBoy said:
Do me a favor and stay out of a tangential discussion if you're not going to address that discussion.

Your point is worthless in the discussion subset me and pxc are having.
Click to expand...

Join Date: Oct 2005
Click to expand...
Ok, which banned phanboi are you? FanATic? If you want to have a private discussion with pxc use PM.

We are discussing ATI's lack of full support for SM3.0 in the R520, as well as a document that they posted where they support VTF (in theory).
 
PRIME1 said:
I have seen no proof that their method will even work. Since it's out of spec developers will have to program a seperate path for it, or ATI will have to translate the instructions with in the driver. That does not sound better to me.
Click to expand...

since you didnt catch it the first time

Part of the point of vertex texturing is to be able to expose pixel format data to the vertex shader, and as somewhat of an alternative to Vertex Texturing ATI will be promoting the use of a new extension to DirectX known as Render to Vertex Buffer. As the name implies Render to Vertex Buffer (R2VB) allows all of the operations within the pixel shader to be utilised, but rather than rendering to a displayable surface or texture the results are rendered to a buffer in memory that can be directly read as an input to the vertex shader. The upshot of this process is that an application can have access to the capabilities of the Pixel Shader which can then be fed back into the geometry processing pipeline, which should result in a superset of capabilities of vertex texturing and should actually perform better than current vertex texturing schemes because the pixel pipelines are inherently built to cope with, and hide texture latencies.
Click to expand...
 
PRIME1 said:
Ok, which banned phanboi are you? FanATic? If you want to have a private discussion with pxc use PM.

We are discussing ATI's lack of full support for SM3.0 in the R520, as well as a document that they posted where they support VTF (in theory).
Click to expand...

Exactly, and we've already discussed it except you seem incapable of reading what others, immensely more knowledgeable than you, have written with regards to it. Yet, you somehow have declared that you are technically proficient enough in this field to disagree with those who have written this specification and have decided that "hack" or not, ATI x1x00 cards are full SM3.0 cards.

Second, your reply to my post again evinces your complete lack of ability to read as I wasn't talking to you nor did your post add anything useful to the discussion other than just show your ignorance on the topic. If you and pxc somehow understand MS's spec better than MS then I would urge you to go ahead and mention it to them. Otherwise, your parroting of a sentence over and over proves nothing other than you don't know when to stop talking about something you just don't understand.

Finally, your pathetic attempt to label me a "phanboi" (hello kettle?) is sad. Resorting to name calling shows your immaturity and that you lack the ability to make a coherent argument to support the position you are espousing.
 
OldBoy said:
blah, blah, blah
Click to expand...
Show me where MS supports the lack of VTF? Show me where any MS representative has commented on this. Show me some proof or fact to back up anything you have said.

Before you go judging my knowledge of this subject, show some of your own.

As for my statement that you found insulting.... All someone has to do is look at your profile and review the hundred+ posts you have managed to put up in just a few short days to see a pattern.

While clearly I don’t agree with everyone in this thread at least they are bringing something to the table, so far you have brought absolutely nothing. You are now on my ignore list because your posts are no better than spam.
 
PRIME1 said:
Show me where MS supports the lack of VTF? Show me where any MS representative has commented on this. Show me some proof or fact to back up anything you have said.

Before you go judging my knowledge of this subject, show some of your own.

As for my statement that you found insulting.... All someone has to do is look at your profile and review the hundreds of posts you have managed to put up in just a few short days to see a pattern.

While clearly I don’t agree with everyone in this thread at least they are bringing something to the table, so far you have brought absolutely nothing. You are now on my ignore list because your posts are no better than spam.
Click to expand...

That was mature.
 
PRIME1 said:
Show me where MS supports the lack of VTF?
Click to expand...

run DCT, does it have a VTF test yes/no? you have your answer....now stop acting like a prick..
 
Jima13 said:
run DCT, does it have a VTF test yes/no? you have your answer....now stop acting like a prick..
Click to expand...
Hooray! I went 2 years without putting anyone on my ingnore list and today I get to put 2.

Look at the official SM3.0 spec, is VTF in there? YES. End of story.
 
PRIME1 said:
Hooray! I went 2 years without putting anyone on my ingnore list and today I get to put 2.

Look at the official SM3.0 spec, is VTF in there? YES. End of story.
Click to expand...

does it state in the offical sm3.0 spec that VTF is REQUIRED?

waiting on your answer
 
Digital Viper-X- said:
does it state in the offical sm3.0 spec that VTF is REQUIRED?
Click to expand...
That is a gray area that is being discussed right now. It does not spell out whether it is optional or not. However, it is in the spec and developers programming to the spec are using it. How this will affect either card company in the future will become clearer when those games come to market.
 
PRIME1 said:
That is a gray area that is being discussed right now. It does not spell out whether it is optional or not. However, it is in the spec and developers programming to the spec are using it. How this will affect either card company in the future will become clearer when those games come to market.
Click to expand...


So then since its in a gray area that when you state that VTF is REQUIRED because its not OPTIONAL, you're simply stating your opinion and not fact right?
 
Digital Viper-X- said:
does it state in the offical sm3.0 spec that VTF is REQUIRED?
Click to expand...
Compliance is an all or nothing thing, a rule MS started with DX8. Read the first post of this thread and look at ATI's GDC presentation from a few months ago, page 4.

On the flip side, where does MS say it's optional? I've been asking for 2 days and no one has produced a shred of official or unofficial Microsoft documention stating it's optional. Neither the SDK documentation (search the online MSDN archive if you don't have it installed) nor my Microsoft Press DX9 book say VTF is optional.
 
pxc said:
Compliance is an all or nothing thing, a rule MS started with DX8. Read the first post of this thread and look at ATI's GDC presentation from a few months ago, page 4.

On the flip side, where does MS say it's optional? I've been asking for 2 days and no one has produced a shred of official or unofficial Microsoft documention stating it's optional. Neither the SDK documentation (search the online MSDN archive if you don't have it installed) nor my Microsoft Press DX9 book say VTF is optional.
Click to expand...


Fair enough, but also the Beyond3D article mentions that

"R2VB is actually a subset of ATI's OpenGL "Uberbuffers" extensions (but a superset of current Vertex Texturing methods)"

maybe my english is bad but wouldnt R2VB being a SuperSet of the current VT Methods mean anything? for example VB uses ASCII where Java uses Unicode, UC Being a superset of ASCII(Ascii is avaiable within unicode?) so wouldnt that techincally mean that the current VT methods are avaiable in R2VB? i dunno im confused 0.0 maybe someone can clear it up
 
Digital Viper-X- said:
maybe my english is bad but wouldnt R2VB being a SuperSet of the current VT Methods mean anything?
Click to expand...
Funny you should use "superset". We wouldn't be having this discussion if ATI supported VTF. So it's not a superset of the VS3.0 spec. R2VB is a Pixel Shader operation. VTF is a VS3.0 operation.
 
PXC is handling this far better than Prime. Prime, ever since the X1Ks were launched you just flaunted your inner F.A.N B.O.Y to the max. You seem to not mind the Shimmering and driver problems with most releases but when ATi has something as little as this you just shoot off about how they lie and are bad people. Jesus man calm down and look at the situation. From what Ive seen most people are claiming this feature is faster via ATi's way anyway. We will have to wait and see and thats that.
 
PRIME1 said:
http://www.ati.com/developer/gdc/D3DTutorial01_3_0_Shaders.pdf

It seems back in March ATI did a presentation at the Game Developers Conference. While talking up the virtues of Vertex Shaders there is no mention of them not including the Vertex Texture Fetch, nor are they instructing anyone about a "work around"

Maybe the R580 or R600 will have this feature. The R520 may have had problems with it so they dropped it.
Click to expand...
Oh my. How in the great, wide world did I see you finding more info about this "issue"?
 
{NG}Fidel said:
PXC is handling this far better than Prime. Prime, ever since the X1Ks were launched you just flaunted your inner F.A.N B.O.Y to the max. You seem to not mind the Shimmering and driver problems with most releases but when ATi has something as little as this you just shoot off about how they lie and are bad people. Jesus man calm down and look at the situation. From what Ive seen most people are claiming this feature is faster via ATi's way anyway. We will have to wait and see and thats that.
Click to expand...
Shimmering was fixed. It was a bug. We are not discussing a bug here. If I recall there was also a shimmering issue with ATI as well.

Should this information be suppressed so that your image of ATI as a pure and perfect corporation can be maintained? I guess I could be like you and go into the thread about NVIDIA shimmering and label everyone a fannyboy.

Perhaps the forums at Rage3d would be more to your tastes.

Why is everyone getting so upset about this? Can't anyone relax and have a debate about the issue?
 
banGerprawN said:
Oh my. How in the great, wide world did I see you finding more info about this "issue"?
Click to expand...
I'm sure in the next couple months I will dig up even more ;)

Info is what we are all about.
 
Prime1, I have 0 interest in getting into any kind of pissing match with you, but if you truly believe you are even remotely objective when it comes to these two graphics companies then the ability for the human mind to deceive itself never ceases to amaze me.
 
pxc said:
Funny you should use "superset". We wouldn't be having this discussion if ATI supported VTF. So it's not a superset of the VS3.0 spec. R2VB is a Pixel Shader operation. VTF is a VS3.0 operation.
Click to expand...

Render to Vertex Buffer looks like its a Vertex operation done through the pixel shaders? =p
 
John Reynolds said:
Prime1, I have 0 interest in getting into any kind of pissing match with you, but if you truly believe you are even remotely objective when it comes to these two graphics companies then the ability for the human mind to deceive itself never ceases to amaze me.
Click to expand...
That's the first problem. This is NOT a pissing match. Is VTF in the spec? Is VTF supported on ATI cards? That's the only issue here. A lot of other issues have been brought up during the thread; however I will honestly say that at this point I don't know what the full aspect of this will be.

It's hard for anyone to remain objective when taking sides in an issue. Point in case, I have received more personal attacks in this thread for posting this issue than I have relevant Reponses backed with fact or knowledge.

If you truly think I am wrong, please quote me and post a fact that disputes what I have said. Most people have been just quoting beyond3d as their source for information. While I agree many of those people there are experts in the field, they did not write the spec nor did they even come to clear conclusion as to what overall affect this will have.
 
Digital Viper-X- said:
Render to Vertex Buffer looks like its a Vertex operation done through the pixel shaders? =p
Click to expand...

Let's see how important quibbling over stuff like this is when the industry moves to a unified shader architecture next fall.

Anyone want to hazard a guess as to just how many PC titles will make use of VTF before R600/NV50 are released next year? We've had almost 1.5 years of SM3.0 hardware availability so far with one title making use of this feature (Pacific Fighters), so will that trend continue or is this "major" feature on the cusp of receiving widespread developer support?
 
repeat after me: the only people it matters to are forum video advocates.

I have a gtx and I couldn't care less---note proper use of the couldn't care less phrase---about the fact that the ati part doesn't support some bit of sm3.0.
 
PRIME1 said:
Shimmering was fixed. It was a bug. We are not discussing a bug here. If I recall there was also a shimmering issue with ATI as well.

Should this information be suppressed so that your image of ATI as a pure and perfect corporation can be maintained? I guess I could be like you and go into the thread about NVIDIA shimmering and label everyone a fannyboy.

Perhaps the forums at Rage3d would be more to your tastes.

Why is everyone getting so upset about this? Can't anyone relax and have a debate about the issue?
Click to expand...

OMG, are you seriously claiming "everyone else" is getting upset but you are somehow the voice of reason? As has been said, the ability of the mind to decieve itself is strong in you.
 
Digital Viper-X- said:
offtopic question, in VS3/PS3 specs, is HDR part of the specs?
Click to expand...

Not that I know of. That is why the X800, 6800/7800, and X1800 all have different implementations of HDR.
 
John Reynolds said:
Let's see how important quibbling over stuff like this is when the industry moves to a unified shader architecture next fall.

Anyone want to hazard a guess as to just how many PC titles will make use of VTF before R600/NV50 are released next year? We've had almost 1.5 years of SM3.0 hardware availability so far with one title making use of this feature (Pacific Fighters), so will that trend continue or is this "major" feature on the cusp of receiving widespread developer support?
Click to expand...
So you don't think people have been programming SM3.0 games for the last 1.5 years and you don't think that they will continue to program games using the SM3.0 spec for the next couple years?
 
Status
Not open for further replies.
Back
Top