Register Reports R520 to launch with unified shaders Arch

[Turtle 1]

GOOD one

 

[Turtle 1]

Originally posted by: Gamingphreek
ATI already said, way back when, that the R520 will not be a Unified Shader Architecture... they did however hint that there was a strong possibility for the R600 to be Unified.

As for SM4. Yeah........Ok :roll: You keep wishing that.

This is the most ridiculous thread ever. I feel like someone at a Howard Dean rally. (Turtle) And if it has SM4... Were going to RUB IT IN THERE FACES!!!!

The ignorance is ridiculous right now. (Duvie for mod 🙂 🙂 )

-Kevin

That wasn't very nice

 

[nRollo]

Originally posted by: M0RPH

Originally posted by: Ronin
Turtle, you have some perception that anyone gives a damn what you say (and that includes AT 'management', which you seem to throw around quite heaviliy). We don't need to ask you wwybywb, because we already know.

Stop the plagarism, stop the banter (because you really can't keep up), and take responsibility for your own actions in this thread. All you've achieved here is confirming that you're clueless.

Points of note:
The R520 will not have unified shaders.
~80% of the information you've seen on the net is correct about R520 (based on the information I have, the core is correct, but the memory is not).
The piping architecture is slightly different than previous versions (I'm not talking about memory here).

Wow, we have an industry insider here. Doubtful.

We've all known for a long time that Ronin works in the gaming industry and gets both parts and software before most people can Morph. I've never seen any reason to doubt his information.

 

[nRollo]

Originally posted by: Turtle 1
http://www.theinquirer.net/?article=14050

LOL Teh Turtle.

If you are not Intelia, you are a reasonable facsimile.

 

[Pete]

Originally posted by: Turtle 1
Pete I concur but what exactly is SM3+ As I understand it it is the consol version of SM4
so would that be like DX9L

AFAIK:

SM3+ is SM3 plus some extra features. The GF 6- and 7-series are both SM3+ because FP blends aren't a required part of the spec, yet they support it. Both next-gen console parts are SM3+ (to what degree, I'm not sure).

DX9L is SM2 with some extra bits, not SM3.

As this is a public forum, anyone can respond to anyone else in any way they want, so you'll just have to deal. 🙂

Also, it's just about the worst netiquette to do as you have done and post articles, word for word, without even a link. At the very least include a link to the article, and at best quote just a small part of the article and link the rest (to give the publisher some ad revenue).

Here are the links to the two articles in your quoteless, linkless post (gg Google):
http://www.theinquirer.net/?article=20919
http://www.theinquirer.net/?article=14050

You're employed by The Inq? I find it hard to believe that someone who works for an ostensibly journalistic enterprise doesn't properly attribute sources. Please do so in the future, it's just easier for everyone and polite for the quoted authors.

And that IGN "article" is Microsoft PR, so beware of outsize claims.

M0rph, those are some nice quotes you're providing.

The attraction of unified shaders in hardware is to help alleviate bottlenecks, or to allow devs more creative freedom. Yes, discrete pixel and vertex shaders may allow for more efficient transistor usage (and thus greater shader power per transistor). But unified shaders allow devs to focus more heavily on vertex than pixel shading in certain scenes or even for the whole game, essentially reducing bottlenecks and restraints, respectively.

(Dave's article at Beyond3D on Xenos is as close to required reading for unified shaders as it gets ATM, IMO.)

But unified shaders do not imply SM4 compatability, as Xenos demonstrates.

R520 will not have embedded memory for framebuffers, either, as it's not economically feasible to include enough for typical PC resolutions. I'm pretty sure an ATI exec has been quoted saying as much in the past few months. R600, being a successor to R400, should be ATI's first unified shader GPU for the PC.

 

[Ronin]

If Turtle works for The Inq, perhaps I should call Charlie (yes, I have his #) and see if this is fact or not.

 

[videogames101]

Originally posted by: Ronin
If Turtle works for The Inq, perhaps I should call Charlie (yes, I have his #) and see if this is fact or not.

plz do, i would like to know.

 

[5150Joker]

Turtle works at The Inq? :Q

 

[SickBeast]

Originally posted by: videogames101

Originally posted by: Ronin
If Turtle works for The Inq, perhaps I should call Charlie (yes, I have his #) and see if this is fact or not.

plz do, i would like to know.

There's no way that he does; he can't even spell!

 

[Turtle 1]

Originally posted by: Pete

Originally posted by: Turtle 1
Pete I concur but what exactly is SM3+ As I understand it it is the consol version of SM4
so would that be like DX9L

AFAIK:

SM3+ is SM3 plus some extra features. The GF 6- and 7-series are both SM3+ because FP blends aren't a required part of the spec, yet they support it. Both next-gen console parts are SM3+ (to what degree, I'm not sure).

DX9L is SM2 with some extra bits, not SM3.

As this is a public forum, anyone can respond to anyone else in any way they want, so you'll just have to deal. 🙂

Also, it's just about the worst netiquette to do as you have done and post articles, word for word, without even a link. At the very least include a link to the article, and at best quote just a small part of the article and link the rest (to give the publisher some ad revenue).

Here are the links to the two articles in your quoteless, linkless post (gg Google):
http://www.theinquirer.net/?article=20919
http://www.theinquirer.net/?article=14050

You're employed by The Inq? I find it hard to believe that someone who works for an ostensibly journalistic enterprise doesn't properly attribute sources. Please do so in the future, it's just easier for everyone and polite for the quoted authors.

And that IGN "article" is Microsoft PR, so beware of outsize claims.

M0rph, those are some nice quotes you're providing.

The attraction of unified shaders in hardware is to help alleviate bottlenecks, or to allow devs more creative freedom. Yes, discrete pixel and vertex shaders may allow for more efficient transistor usage (and thus greater shader power per transistor). But unified shaders allow devs to focus more heavily on vertex than pixel shading in certain scenes or even for the whole game, essentially reducing bottlenecks and restraints, respectively.

(Dave's article at Beyond3D on Xenos is as close to required reading for unified shaders as it gets ATM, IMO.)

But unified shaders do not imply SM4 compatability, as Xenos demonstrates.

R520 will not have embedded memory for framebuffers, either, as it's not economically feasible to include enough for typical PC resolutions. I'm pretty sure an ATI exec has been quoted saying as much in the past few months. R600, being a successor to R400, should be ATI's first unified shader GPU for the PC.

Thanks pete And the unauthored artlce was presented as it was there was no author .

Ya I read more about the embedded memory. I got the same thing you did after reading ATI ece. artcle. Can you tell me about the 300-350 million transistors rumor in R520.

 

[Turtle 1]

First I am this woman Intelia now I work for the Inquirer LOL

 

[nRollo]

Originally posted by: SickBeast

Originally posted by: videogames101

Originally posted by: Ronin
If Turtle works for The Inq, perhaps I should call Charlie (yes, I have his #) and see if this is fact or not.

plz do, i would like to know.

There's no way that he does; he can't even spell!

Do not doubt teh Turtle's powers Sickbeast! He has come from the future to tell us of a time when ATIs patent from 2003 ends all computer gaming as we know it.

He is like Kyle in "Terminator 2"- he comes from a time where nVidia does not exist because they made a comment about unified shaders a few years ago.

Teh Turtle is a wise visionary who had no time to learn to spell, only to build a time machine and come warn us to make the right video card choices now, because in two years none of us will be able to buy Vista compliant hardware for reasons as yet unknown to us.

Teh Turtles posts are too advanced and scientific for us to comprehend, we can only do as he says and pray.

🙁

 

[5150Joker]

Hey to go a little OT here, did anyone notice a performance decrease in BF2 with the 81.26 drivers?

 

[Keysplayr]

Originally posted by: Turtle 1
At no time do I take credit for authoring that article . It comes from my employers archives and have full permission to use it.



http://xbox360.ign.com/articles/617/617951p3.html

Here it is.

 

[Keysplayr]

Originally posted by: Turtle 1
First I am this woman Intelia now I work for the Inquirer LOL

Yes, you said you quoted directly from your employer's archives and had full permission to do so. Did you not?

 

[Gamingphreek]

Kelvin beings how its all old news to you...cough Explain what embedded memory is going to do for us gamers.

Ok are you seriously that stupid. Really, i posted the answer a couple of posts up!!! Not only that how in the hell do you get my name wrong. It is only 5 letters and is at the end of EVERY one of my posts.

#1. Kevin not Kelvin.

#2. Embedded DRAM, is so ATI can, essentially, enable 4x AA for free. It is not on the R520 however there is very strong evidence that it is in the XBOX 360.

Duvie I have not flamed 1 person . Your constant acrimonious statements .Will not go unnoticed by the management of Anandtech. The fact there has been flame here is true but none by me.

HAHAHAHAHAHHA. Omg this guy must be like 12 years old. Seriously stop Inte...i mean Turtle... really im laughing way too much.

Wow, we have an industry insider here. Doubtful.

Everyone reading this, please note that the above statements are OPINIONS, not FACT.

I don't believe that the R520 will have unified shaders either but I would never have the gall to state is as fact when I don't know the specs yet since they haven't been announced.

You must be in the same grade Turtle is. Everything Ronin made was a direct statement. There is no way you can turn that into opinion. He doesn't even take a stance. He merely lists off some stuff he knows.

Additionally, he does work for a software company. Doubt it? Go look at his other posts. Go to the Nvidia forums. Even you cant be stupid enough to overlook some of them.

Yes, you said you quoted directly from your employer's archives and had full permission to do so. Did you not?

Caught in a bit of a pickle there aren't ya Turtle.

-Kevin

 

[Gstanfor]

Originally posted by: Turtle 1
Thats correct. and you guys think Nvidia is going to pull a rabbit out of its hat ?its also a repeat of the unauthored text. So you new that Fast14 was part of the new R500/R520 GPU'S But not everyone else did.


Speculation only

LOL!

nVidia has long since pulled the optimised math logic bunny out of its hat, Turtle 1 - you really should try to keep up.

ATi bas been rabbiting on about Fast14 for years now, meanwhile nVidia has been quietly taking advantage of Arthmatica's CellMath technology since the nV4x series lauched!

http://www.arithmatica.com/customers_1.html

 

[Turtle 1]

Originally posted by: keysplayr2003

Originally posted by: Turtle 1
First I am this woman Intelia now I work for the Inquirer LOL

Yes, you said you quoted directly from your employer's archives and had full permission to do so. Did you not?

Yes Keys thats what I said . But thats not were I pulled that article from.

 

[Pete]

Weird, I think the forum ate my post. I thought I replied after keys and before Kevin, three posts up.

Anyway, NP, Turtle. No idea about the R520 transistor count, but not much longer to wait.

I assumed you worked at The Inq b/c you said you had your employer's permission to post those two articles, and they're both The Inq's. Those are the unattributed reprints everyone's hounding you about. (I suspected they were from The Inq or The Reg b/c of the writing style, and Google found the links right quick.)

The unauthored article, that PR reprint at IGN, was actually posted by a Major Nelson on his blog, IIRC. But it's MS PR, and the IGN link made that clear, so there was no real need for you to specify the author, as you provided a link.

Ronin, I'd only be impressed if you had Fuad's #. 😀

 

[Turtle 1]

Originally posted by: Gstanfor

Originally posted by: Turtle 1
Thats correct. and you guys think Nvidia is going to pull a rabbit out of its hat ?its also a repeat of the unauthored text. So you new that Fast14 was part of the new R500/R520 GPU'S But not everyone else did.


Speculation only

LOL!

nVidia has long since pulled the optimised math logic bunny out of its hat, Turtle 1 - you really should try to keep up.

ATi bas been rabbiting on about Fast14 for years now, meanwhile nVidia has been quietly taking advantage of Arthmatica's CellMath technology since the nV4x series lauched!

http://www.arithmatica.com/customers_1.html

If your referring to the performance of a single card 6800 Ultra vs. a single X850xt I seen no rabbit come out of that hat.

If your referring to a G70 vs. XT850xtpe I still see no rabbit .

Now if the G70 out performs a R520 I will see a rabbit . But for another 10 days I see no rabbit

 

[Turtle 1]

Ever since I bought TI 4600 I have ignored nvidia.

 

[Turtle 1]

Thanks Pete as I didn't know the source I just pulled it from the Data file./ I will add that stuff to that article so as not to make that mistake again. You saved me some time . I still don't no why it was in the file with out the sources. shabby work by me . probably a long day by me.

 

[crazydingo]

Originally posted by: Pete
Anyway, NP, Turtle. No idea about the R520 transistor count, but not much longer to wait.

My guess: 230-240 million.

 

[Turtle 1]

Thats about what R500 has without the embedded memory. I would think thats low however. The rumors have been 300-350 million . We all have to admit 1 thing. Everyone can agree to.The R520 has had more rumors than anything I have seen in a long time.

Vista has been in the making along time and I think R520 beats it in the rumor mill.

 

[xtknight]

Originally posted by: crazydingo

Originally posted by: Pete
Anyway, NP, Turtle. No idea about the R520 transistor count, but not much longer to wait.

My guess: 230-240 million.

Maybe I'm smoking crack, but wasn't the GeForce 6800 Ultra 300 million transvestites?

Edit: never mind, it was 222 mil.

My guess is with AVIVO and all, the R520 will be 300 million.