R420 Might Support Branching after all.

[KimB]

It seems that finally the ATI R420 will have Directx 10 features. Does this involve that it support VS 3.0 and some VS 4.0 elements ????

More likely it's related not to VS/PS support, but rather something different, such as HOS support. Either way it seems doubtful, but fortunately we don't have to wait long to find out.

 

[BRiT]

thats assuming its for the r420 .

It could be for any other card

That's the key. What other card could the PS 2_b profile be for? Well, only about a week to wait.

I'll say it again... R350/R360 and RV370/RV380 ?

 

[DemoCoder]

How do you resolve that with the Huddy presentation, and the various "talking down" messages of PS3.0 of various ATI employees on B3D?

 

[Frank]

How do you resolve that with the Huddy presentation, and the various "talking down" messages of PS3.0 of various ATI employees on B3D?

Well, that might make sense if the R420 is actually (more or less) a DX10 part, wouldn't it?



Damn, one more week to wait before we know for sure.

 

[Trawler]

How do you resolve that with the Huddy presentation, and the various "talking down" messages of PS3.0 of various ATI employees on B3D?

It's just ATI spreading FUD, as usual.

 

[hoom]

I've only seen them hosing down dynamic branching & only in the 'hurts performance if not done very carefully' way...

Anyways, I believe the 'dx10 stuff' bit is just a mistranslation.

 

[BRiT]

I've only seen them hosing down dynamic branching & only in the 'hurts performance if not done very carefully' way...

Anyways, I believe the 'dx10 stuff' bit is just a mistranslation.

The DirectX 10 thing is in there exactly the same in the native (foreign language) page. It is not a mistranslation by babelfish/googlefish.

 

[ninelven]

Just say no...

 

[KimB]

I'll say it again... R350/R360 and RV370/RV380 ?

It doesn't make any sense that the PS 2_b profile is a R3xx chip + F-buffer. The PS 2_b profile has the same dependent texture instruction limit as plain PS 2.0. Using the F-buffer would shatter this limit.

 

[BRiT]

I'll say it again... R350/R360 and RV370/RV380 ?

It doesn't make any sense that the PS 2_b profile is a R3xx chip + F-buffer. The PS 2_b profile has the same dependent texture instruction limit as plain PS 2.0. Using the F-buffer would shatter this limit.

So why would it make sense for the R420?

 

[KimB]

I'll say it again... R350/R360 and RV370/RV380 ?

It doesn't make any sense that the PS 2_b profile is a R3xx chip + F-buffer. The PS 2_b profile has the same dependent texture instruction limit as plain PS 2.0. Using the F-buffer would shatter this limit.

So why would it make sense for the R420?

I don't think you understood me. I said that the PS 2_b profile just can't be one of ATI's current chips with F-buffer support finally enabled (due to the reason above....). This doesn't have anything to do with the R420.

 

[jvd]

I'll say it again... R350/R360 and RV370/RV380 ?

It doesn't make any sense that the PS 2_b profile is a R3xx chip + F-buffer. The PS 2_b profile has the same dependent texture instruction limit as plain PS 2.0. Using the F-buffer would shatter this limit.

So why would it make sense for the R420?

I don't think you understood me. I said that the PS 2_b profile just can't be one of ATI's current chips with F-buffer support finally enabled (due to the reason above....). This doesn't have anything to do with the R420.

no bug it could be a refresh of a current ati card that has a few options to make it that way .

 

[christoph]

in rendermonkeys latest version there is no 3.0 support either....

 

[Bry]

I'll say it again... R350/R360 and RV370/RV380 ?

It doesn't make any sense that the PS 2_b profile is a R3xx chip + F-buffer. The PS 2_b profile has the same dependent texture instruction limit as plain PS 2.0. Using the F-buffer would shatter this limit.

So why would it make sense for the R420?

I don't think you understood me. I said that the PS 2_b profile just can't be one of ATI's current chips with F-buffer support finally enabled (due to the reason above....). This doesn't have anything to do with the R420.

Using your train of thought and logic DX9.0b would be better suited for the 9800XT than it would for R420 then..The R420 should shatter the 9800XT's limits It is quite reasonable that DX9.0b is being released as an upgrade for R3XX and for the replacements to the 9600 series of cards. Yet the X800 and X800XT is fully SM3.0 compliant..ATI showed with the 9100 that they try not to released cards and say it is compliant when it can't do it (AKA 5200) So instead of lying and saying these lower end (and soon to be last gen) cards are fully complaint..They can enable them to do what they can do without misleading the public. Just a thought

 

[Scali]

I suppose you meant DX9.0c?

 

[Xmas]

Why is it that, despite all indication, this rumour still exists?

 

[jvd]

Why is it that, despite all indication, this rumour still exists?

lots of nvidia fans spreading it ?

It could very well be true. But it makes almost no sense the way people are putting it together

 

[Xmas]

Why would NVidia fans spread the rumour of R420 supporting SM3.0?

 

[London Geezer]

Why would NVidia fans spread the rumour of R420 supporting SM3.0?

Was wondering the same thing....

 

[dksuiko]

Why would NVidia fans spread the rumour of R420 supporting SM3.0?

Because if you're going to blame anybody for anything.. blame nVidia or its fans.