ATI XBOX 2 etc etc performance etc etc

[fxtech]

from chipzilla , full of details , confirmed edram frame buffer and some performance details worth of read. ...and bye bye to directx 10 i think no VS and PS 4.0

----------------------------------------------------------------------------------

THE NEXT GENERATION Xbox chip will be ready as soon as the end of the first quarter of next year...
This chip comes as a derivatove of the R520 chip better known under its code name Fudo and will be of a similar architecture to the upcoming desktop chip. The only main difference will be an integrated frame buffer that you don’t see that often on desktop cards.

At this time, sources said the Xbox 2 development kit includes a 9800XT card and platform and you can only achieve five to 10 FPS with Xbox 2 working titles.

That’s why a full developer's kit will be available in late Q1 2005. This kit will include the flagship R500 chip that will give the right measure of graphics performance.

From an aarchitectural point of view, we are talking about at least 16 pipelines powered by two TMUs (Texture Memory Units) which roughly give you twice as much power as the current X800 generation of cards that have only one TMU. This will occur only in multitexture games but nowadays all games use more then one texture per pass.

If Xbox games works at 60 FPS you can imagine that R500 will end up more then six times faster then Radeon 9800. R500 looks like a heck of a good chip to me so does Fudo - R520 I guess. µ


EDITED my comment on top

 

[bbot]

Looks like Dave Baumann is wrong after all. XBox 2 vpu will be based on R520 (Fudo), not on the cancelled R400. No Shader Model 4.0, boo hoo!!!!

 

[thop]

I'm kinda dissapointed. If PS3 will be what it is rumoured to be it is gonna smoke that thing.

 

[London Geezer]

And i should believe anything that guy said because..............................?

 

[Ender]

Hmm..
I don´t know about this. I´ve been hearing that the WGF, the Windows Graphical Foundation (the next DirectX) will use Shader 4.0 and that Xbox2 will support that aswell...

 

[speng]

umm references???

Speng.

 

[London Geezer]

Seen how WGF is coming out relatively shortly, i think Xbox2 will at least be SM4. If not more..

 

[Evil_Cloud]

Too many gramatical errors to believe this part of news.

 

[fxtech]

slow down , i only repoted a news from www.chipzilla.org.

i hope deanoc will have time to spend some word about it

 

[one]

slow down , i only repoted a news from www.chipzilla.org..

i hope deanoc will have time to spend some word about it

Chipzilla or, more friendly, The Inquirer.

Oh, and now Microsoft employees using Xenon moniker in public, maybe the "leaked" spec was intentionally posted by the team.

 

[Inane_Dork]

From an aarchitectural point of view, we are talking about at least 16 pipelines powered by two TMUs (Texture Memory Units) which roughly give you twice as much power as the current X800 generation of cards that have only one TMU. This will occur only in multitexture games but nowadays all games use more then one texture per pass.

Is there really a good reason to go from a 16x1 to a 16x2 pipe design? Sure, it may go faster, but given how math ops are becoming more prominent, I doubt this is the right move.

If Xbox games works at 60 FPS you can imagine that R500 will end up more then six times faster then Radeon 9800.

60 FPS Xbox games has literally nothing to do with how much faster the R500 will be than the 9800. Was he trippin' when he typed this or...?

 

[DeanoC]

If I laugh any harder I'll wet myself.

 

[Brimstone]

I don't get it. If the GPU does turn out to be a variant of a upcomming desktop chip, why did Nvidia make a big stink about the magnitude of the X-Box 2 contract?

 

[chrisrt2]

I'm kinda dissapointed. If PS3 will be what it is rumoured to be it is gonna smoke that thing.

Any we know exactly what the PS3 is going to compose of and perform like? People, we wont know who is better 'likely' for a few months before launch...

 

[hey69]

The Xenon lives, thanks to a job posting on Microsoft’s website, with an advert for a software engineer confirming the development name for the Xbox successor for the first time.

“Want to help create a new feature for Xenon not available on any other platform? Microsoft Game Studios Publishing is looking for an SDET with server experience to…†Cool.

And there’s more dumplings in the stew, with all manner of revelations emerging. Check this – the successful applicant will create, “a spectator mode for many Xenon games, Xbox TV - with tickers at the bottom of the screen featuring recent high scores and game highlights, a tournament system, and tradable trophies.â€

Meaning that the Xbox 2, or Xenon, or whatever it will be called when it finally rolls onto shelves, will have a load of cool online features on an uprated Live service.

We contacted Microsoft this morning for clarification on the somewhat mysterious Xbox TV mention in the advert, but received a flat 'no comment' off the bat. Indeed, SPOnG believes that someone at Microsoft’s recruitment seriously dropped the ball in posting the information available to the public, given the startled reaction of the firm’s press office.

To reveal a massive range of new features at this stage of the R&D of the planned Xbox Live overhaul, to the public and, more pertinently, to Sony Computer Entertainment, will no doubt result in a certain amount of panicked shouting at Microsoft Towers.

We are scheduled to be contacted with an official comment on exactly what Xbox TV is later today. If we get anything more than a ‘no comment’ we’ll let you know.


have a Spong Day ..

 

[Dave Baumann]

Looks like Dave Baumann is wrong after all.

Yeah, right.

 

[jvd]

As far as i know this is wrong.

He even has the r520 wrong. That would be 24x1 fp32 strait through (no partial persion) and shader model 3.0

as for the r500 , i dunno if its sm 4.0 ( i don't think so last i heard it was more like 3.8 ) but it will be much more powerfull than the r520 in shading

 

[Dave Baumann]

He even has the r520 wrong. That would be 24x1 fp32 strait through (no partial persion) and shader model 3.0

No.

 

[jvd]

He even has the r520 wrong. That would be 24x1 fp32 strait through (no partial persion) and shader model 3.0

No.

No on what dave the fp32 ? (that be the only thing i think is wrong)

 

[Inane_Dork]

He even has the r520 wrong. That would be 24x1 fp32 strait through (no partial persion) and shader model 3.0

No.

Nice. In the matrix of 548 possibilities for R520 specs, one box is now impossible. Only 546 to go.