ATi R600

[no-X]

Srovnání počtu pinů R600 a G80:

R600...



zelené = 20*20 = 400
modré = 18*11*4= 792
růžové = ((41*2)+5)*4 = 348
oranžové = 47*3*4 = 564
rohy = 9*4 = 36
------------------------
celkem: 2140


G80:



oranžová: 2*41*4-4=324
světlemodrá: 2*33*4=264
světlezelená: 13*13=169
červená: 2*4*25=200
modrá: 17*4=68
zelená: 21*2*4=168
žlutá: 4*4*4=64
růžová: 8*4*4=128
šedá: 8*2*4=64
------------------
celkem: 1449

[Masster]

It seems that HEXUS are on a roll with speculation regarding AMD's next-generation R600 GPU, with no less than four stories regarding the part - Some of it outlandish, some of it slightly more plausible. Here's a recap of what they're saying.

R600 has had several challenges, principally over power and cooling.

ATi has looked at water-cooling and traditional air cooling. But, the speed of the processor and its resultant heat output have meant that ATi also had to consider more complex hi-tech solutions.

So, the company turned to NanoFoil cooling technology. This, as best we can understand, works by having ultra-thin nanolayers of aluminium and nickel that can be kicked into life by heat, electricity or mechanical or optical stimulation, causing a reaction that gives off heat in a controlled fashion.

All the might of the AMD/ATi combine can't deliver R600 on time

The reasons why a re-spin might be need can be highly technically convoluted. They can also be ultra simple. Humans, remember, are involved in the production decisions.

With A11, as we understand it, the issue was very, very simple. Someone forgot to connect the pins that let the GPU communicate with the outside world.

ATi keeps on spinning to try to get R600 right

Seemingly, ATi has been aiming to achieve 11,000 3DMark scores at 750MHz. That's tasty and may mean the R600 offers a strong challenge to NVIDIA's G80 GPU given that, as we reported in HEXUS.Beanz a few hours ago, ATi's already got samples running at 1GHz and has higher targets in mind.

Will R600's performance be enough to put ATi back in the lead?

ATi is targeting February 14 - Valentine's Day - for getting R600 review samples into the hands of the press.

And, just in case you were under any doubts. Yes, we do mean next Valentine's Day, not the year after next.

ATi targets CeBIT for R600 launch

So, there you have it - Nanofoil cooling and a 2GHz clock speed, and a Valentine's Day launch to press? Personally, I'd bet against at least two out of those three points being wrong, but who knows?

Opravdu si nekdo mysli, ze je ATI schopna neco podobneho vypustit pristi rok nebo uz opravdu nektere weby uveri kazde kravine?

[no-X]

Když se objevily informace o G80 (128 unified shader units, clock-speed scaleable upto 1,5GHz, 384bit mem.bus), tak to taky znělo šíleně a hlavně kvůli frekvenci "až 1,5GHz" to nebral nikdo vážně.

Pokud je nVidia schopna s 90nm procesem dosáhnout až 1500MHz, tak proč by s 80nm procesem ATi neměla být schopna dosáhnout až 2000MHz? Je sice pravda, že nVidia na to použila bývalé inženýry z AMD, ale pokud budeme brát v potaz, že ATi má k dispozici technologii FAST14, kterou si před časem licencovala, je vcelku možné, že o ní měla zájem právě kvůli R600.

Můj názor je takový, že to sice zní velice šíleně a nepravděpodobně, ale nedá se říct, že je to nemožný. To jsme si mysleli i o některých parametrech G80, nebo o některých (nyní již téměř potrvzených) parametrech R600. Podle mě je velmi pravděpodobné, že některé z uvedených informací pravdivé jsou a jiné ne - zkrátka že je to tradiční směska pravdy a spekulací.

[no-X]

Myslím, že 512bit sběrnici u R600 můžeme považovat za jistotu >:}

R600 has 512-bit external memory bus

Speaking at yesterday's AMD Financial Analysts Day, Executive Vice President of Visual and Media Businesses, Dave Orton, appeared to throw down the performance gauntlet in favor of AMD's upcoming R600 GPU. Having had over a month to study NVIDIA's G80, Orton did not seem the least bit intimidated. In a slide entitled "R600: Why we lead in graphics", Orton promised that even if the name of the company had changed, that the commitment to GPU performance leadership had not. He promised a "take no prisoners" approach to performance leadership for AMD's new GPU.

More interestingly, in his verbal remarks Orton reported (at roughly the 1:22:30 mark of the webcast) that one of R600's key advantages would be "new levels of memory bandwidth to the graphics subsystem, and bandwidth is critical to graphics performance." As all graphics geeks know, AMD pioneered the move to GDDR4 memory with the Radeon X1950 XTX, which gave them a temporary advantage in bandwidth. However, in the period since NVIDIA has released the 384-bit GeForce 8800 GTX, whose memory bandwidth crushes the X1950 XTX by 86.4GB/s to 64.0GB/s. It is impossible that AMD could regain a significant enough advantage in bandwidth to be cited by Orton as a major competitive advantage without following NVIDIA north of the 256-bit bus that has been a mainstay of the ATI/AMD high-end products since 2002's Radeon 9700 Pro.

R600 will therefore feature a 512-bit external memory bus, likely using 1.2GHz GDDR4 for ~153GB/sec from memory pool to chip, to back up the smack AMD's Executive Vice President laid down.

In other tidbits, Orton also vowed to be first to the 65nm technology process, but did not disclose which product he had in mind for the honor, nor even product type, GPU or chipset. Our graphics-oriented notes (and a few selected slides) on the rest of the conference are included inside, if you dare to take the red pill.

http://www.beyond3d.com/

[Adisis]

už se vi kolik karet bude a kolik budou stat?

[no-X]

Pokusil jsem se o shrnutí těch věrohodnějších informací a dohledání původu každé z nich:

- R600 (R.Huddy: “The R600 will be [absolutely] the fastest DirectX 9 chip that we had ever built”)
- 80nm ()
- 96 unifikovaných jednotek (D.Orton: "today it's 48, in the future maybe it's 96")
- 24-32 TMUs (E.Demers: "I would not mind more texture power, but I would not want to reduce the ratio of ALU : TEX")
- 24-32 ROPs (E.Demers: "Right now, they seem to balance at 1:1 (TEX:ROP), but the trend is towards lowering ROPs, in general"
- 512bit (nVidia: "prepositioning against 512bit products", D.Orton: "new levels of memory bandwidth to the graphics subsystem", D.Baumann: "if core speeds scale at a similar ratio at memory, why would we expect any more of the elements that consume bandwidth?"
- 750MHz core ()
- 2,4GHz GDDR4 ()
- 720MTr (>500M... D.Orton?)

[DRTiCZ]

a bude papkat alespon 15A na 12V vetvi

[Adisis]

Pokusil jsem se o shrnutí těch věrohodnějších informací a dohledání původu každé z nich:

- R600 (R.Huddy: “The R600 will be [absolutely] the fastest DirectX 9 chip that we had ever built”)
- 80nm ()
- 96 unifikovaných jednotek (D.Orton: "today it's 48, in the future maybe it's 96")
- 24-32 TMUs (E.Demers: "I would not mind more texture power, but I would not want to reduce the ratio of ALU : TEX")
- 24-32 ROPs (E.Demers: "Right now, they seem to balance at 1:1 (TEX:ROP), but the trend is towards lowering ROPs, in general"
- 512bit (nVidia: "prepositioning against 512bit products", D.Orton: "new levels of memory bandwidth to the graphics subsystem", D.Baumann: "if core speeds scale at a similar ratio at memory, why would we expect any more of the elements that consume bandwidth?"
- 750MHz core ()
- 2,4GHz GDDR4 ()
- 720MTr (>500M... D.Orton?)

diky

[Spajdr]

takže jednotek méně než GTX, ale rychlejší paměti? tzn. větší rozlišení a AA s menší ztrátou oproti G80?

[no-X]

Spajdr: nic zatím není úplně potvrzeno, ale jednotky R600 (pokud nepoběží na nějaké vyšší frekvenci) imho nebudou skalární, ale kombinované, tzn. vektorová pro práci se 3-4 složkami + skalární pro práci s jednou složkou. G80 má skalární jednotky, tzn. mohou pracovat jen s jednou složkou... prostě čistě počtem jednotek se to nedá porovnávat - zítra sem nahodím nějaký teoretický čísla, na jejichž základě už se nějaké závěry dělat dají.

[Spajdr]

Spajdr: nic zatím není úplně potvrzeno, ale jednotky R600 (pokud nepoběží na nějaké vyšší frekvenci) imho nebudou skalární, ale kombinované, tzn. vektorová pro práci se 3-4 složkami + skalární pro práci s jednou složkou. G80 má skalární jednotky, tzn. mohou pracovat jen s jednou složkou... prostě čistě počtem jednotek se to nedá porovnávat - zítra sem nahodím nějaký teoretický čísla, na jejichž základě už se nějaké závěry dělat dají.

Ok, doufám že ten rozdíl mezi R600 a G80 bude pro NVIDII hodně bolavý

[VOiD]

Ok, doufám že ten rozdíl mezi R600 a G80 bude pro NVIDII hodně bolavý

Zatím to vypadá, že R600 má něco navrch.
http://level505.com/2006/12/30/the-full-ati-r600-test/

[mraky]

No jestli se tomu da alespon trochu verit, tak ty testy dokazuji pomerne nevyladene ovladace.

Ale vysledky ukazuji, ze ti co hraji 3d mark nebudou mit duvod upgradovat

[no-X]

Prý je to fake... není tam nic, co by dokazovalo, že tu kartu měli, navíc ten web byl založen předevčírem, takže asi tak...

[Haste]

Fake to zrejme bude, ale ty hodnoty nevypadaji zas tak nerealne, vlastne mi pripadaji celkem realne ...

[no-X]

To jo, ale hodnoty, který budou o pár procent vyšší, než u GTX, si umí vymyslet každý

[Haste]

To jo, ale hodnoty, který budou o pár procent vyšší, než u GTX, si umí vymyslet každý

To jo no, ale vtipne by bylo, kdyby se fakt trefil, protoze tam nema zadny vyrazny ustrel. Asi by to byla "trefa roku"

[THX]

tak, ze vraj

Update - Authenticity Of This Article

This article is authentic. Our R600 sample is an RTM sample used in the MS CERT process for driver validation (especially for Vista). We are not publishing pictures of the card itself right now, as the card contains major ID tags that we can not remove yet for source protection. We will add pictures of the card itself once we can remove these tags.

je to hned na 1. stranke hned na zaciatku

[Bob van Sikorski]

no takovejhle odstavecek umim napsat taky ... a este k nemu pridat "nezavislou realnou recenzi" R600. pracicka tak na 3 hodiny. jen si tak rikam, kolik mu tahle publicita vydela skrze adsense (zeby vice nez tomu-jehoz-jmeno-se-nevyslovuje? )

[no-X]

Level505....

Update - Site, Upcoming 2560×1600 Benchmark, Pictures

Happy new year! We are sorry that the site slowed down a bit. However, here are the great news: tomorrow we will be testing 2560×1600, and we will be able to provide limited pictures of the card approx. on January 5th/6th already. For everyone to understand why we cannot just “black out” the ID Tags: The major tag is a fine punched hologram foil wrap which cannot be taken off because it is glued onto the front metal around the DVI ports. Taking off the fan is impossible without removing it. We will definitely keep you updated, and appreciate all comments we have received so far.

[Haste]

Heh porad mi prijde, ze tu kartu nemeli v ruce. Jinak by nahodili aspon nejake fotky, klidne i s chladicem proste tak jak jim prisla a napr. v Photoshopu by TAGy jiste sly pekne zacernit, aspon by to vypadalo verohodnej nez jen remcat okolo...

[Masster]

AMD's R600 to suffer another delay

THE FIRST AMD/ATI graphics chip just cannot enter the world without much screaming and wriggling.

In fact, it's looking more like Nvidia is going to be to the market first with an 80 nanometre version of the G80 graphics chip, rather than ATI's 720 million transistor monster.

The last revision of the prototype chip - upon which a certain "pre-review" is based - also suffers from problems which are serious enough to get another re-spin, sources tell me. This re-spin puts a hold on the launch for another couple of weeks, and now R600 is looking like an early March launch, probably the week before SnowBIT in Hangover. However, AMD/ATI is making severe changes to the whole line up and we can say that this launch, when it happens, will be very, very shocking for the 3D industry.

http://www.theinquirer.net/default.aspx?article=36673

[Semik]

Takze problemy s vydanim novych cipov u AMD sa ted prenieslo aj do Ati. To my pripada dost zle, kedze grafiky oproti CPU starnu odost rychlejsie (moralne).
Ak nebude vykon zas navyseny aspon o 10-15% to nebude prave 80nm G80 killer
Aspon by to uz mohlo vyjst aj s novym Ati mainstreamom!

[no-X]

ATi předvede R600 na CES?

AMD shows off two R600 boards at CES

WELCOME TO THE FIRST Edition of INQUIRER's show information bulletin for journalists and bloggers who think they're important enough.

We figured that less influential sites would like to know that AMD has two R600 prototype boards on the show.

We're talking about full-size, 12-inch boards with nice R600 XTX GPU and sixteen Samsung 32MB memory chips - that's right, whole Gigabutt of video memory connected to the GPU via that rumoured 512-bit memory interface. Or at least, there are some people in this industry who consider it as a rumour.

The GPU featured on the boards is not the final one (as we already disclosed), but this will be the first public showing of the product to journalists who are important enough. That is, unless you're a short Italian or tall Croatian and already saw them five months ago.

Regarding the CES boards, we haven't seen them, but look out later for our R600 line-up story. We know that the Green Goblin's minions back in Santa Clara, CA are watching this space with the utmost attention. µ

http://www.theinquirer.net/default.aspx?article=36757

[Masster]

Latest ATI "R600," "R610" and "R630" Details

The R600, ATI's ultra-high-end Radeon X1950 successor, has a production date scheduled for February 2007. The card will launch at or around the Cebit 2007 convention in mid-March. Shipments will follow shortly after.

Our latest roadmaps indicate R600 will support unified shaders and GDDR4 or GDDR3 memory. GDDR3 versions of the card running revision "A12" silicon appear to be making rounds in the blogsphere, and select press can even take a sneak peak of the card under embargo here at CES. The final silicon for R600 will be "A13."

http://www.dailytech.com/article.aspx?newsid=5630