New news about ATI R600

[ownage]

http://level505.com/2006/12/30/the-f...i-r600-test/1/

My first Thread, hope i didn't post in the wrong section.

The R600 Specs are:

* 64 4-Way SIMD Unified Shaders, 128 Shader Operations/Cycle
* 32 TMUs, 16 ROPs
* 512 bit Memory Controller, full 32 bit per chip connection
* GDDR3 at 900 MHz clock speed (January)
* GDDR4 at 1.1 GHz clock speed (March, revised edition)
* Total bandwidth 115 GB/s on GDDR3
* Total bandwidth 140 GB/s on GDDR4
* Consumer memory support 1024 MB
* DX10 full compatibility with draft DX10.1 vendor-specific cap removal (unified programming)
* 32FP internal processing
* Hardware support for GPU clustering (any x^2 number, not limited to Dual or Quad-GPU)
* Hardware DVI-HDCP support (High Definition Copy Protocol)
* Hardware Quad-DVI output support (Limited to workstation editions)
* 230W TDP PCI-SIG compliant

















I Hope this is reliable information, if this information is correct, then R600 will be a great videocard

[Order]

Looks like a decent performer indeed. Thanks for the info.

[lowfat]

results don't look out of this world either.

[TaPaKaH]

230W TDP on a graphics card?

[C Stat B]

results don't look out of this world either.

Agreed. I was expecting better, though these probably aren't accurate.

[SpikeP]

Looks good in places, but not mind blowing. My question would be what is the core clock on the card, and what is the r600 like under overclocking. 8800GTX's are exactly shaby under normal air and normal volts!

- Paul

Okay looking further anno shows a big difference. wonderhow much their is left in driver optimisation as well

[DilTech]

There's something very odd with their numbers for the 8800GTX...

1.) CF x1950xtx at 1600x1200 doesn't speed past the 8800GTX that much. ~10 fps, but damn well not 20. Mind you, the test you've shown us runs a faster processor than the link I'll post, so the framerate should've gone up if anything for the 8800GTX, not down.
http://www.anandtech.com/video/showdoc.aspx?i=2870&p=22

2.) Oblivion runs MUCH faster than those numbers at 1600x1200 with 4xAA than 19 fps on the 8800gtx.
http://www.vr-zone.com/index.php?i=4216&s=22

There's other weird ones(3dmark05 is rather suspect, as at stock with a slower proc I've scored about 1k higher with 8xAA), but I'll give them the benefit of the doubt.

[couppi]

Shens?

No pics of the card.
Site was made 2 days ago.
Strange results sometimes.

Still, it might be legit. Some people have some crazy early access to things.

[cky2k6]

this aint exactly anandtech, so these results are just indicative of that the r600 is a true competitor against g80. i would hope that the results would be better with a mature driver, as nvidia will probably have an answer to r600, and if the r600 is only marginally faster, with some results questionable as well, i dont like ati's prospects. yet to see dx10 performance though, so that should be interesting.

[DilTech]

Thought I'd point this out guys...

Support for DX10.1 draft?

We all know this by now, but incase you don't let me give you a refresher course. Microsoft stated you either 100% match DX10 specifications(this includes all updates) or your card is marked by dx as not compatible... This means ATi is banking on what they expect to be DX10.1 to actually be it to a T.

[Lightman]

Looks good in places, but not mind blowing. My question would be what is the core clock on the card, and what is the r600 like under overclocking. 8800GTX's are exactly shaby under normal air and normal volts!

- Paul

Okay looking further anno shows a big difference. wonderhow much their is left in driver optimisation as well

From simple calculation clock speed seems to be 820MHz .

missing data for evaluation of clock speed:

Processing Power:

* 105,000,000,000 SOps (Shader Operations Per Second)
* 2,000,000,000 TCps (Triangle Calculations Per Second)
* 28 GTps (Gigatexels Per Second)

[oqvist]

because it´s only a small difference in speed and the upcoming version may be more expensive...

You can always wait two month and get a faster video card then if it´s not ATI that has it out it´s nvidia

[phi|os]

R600 is better than these benches.

[Ubermann]

They created a new site just to show this ?
Hmm..

[cky2k6]

i guess its a bigger site that didnt want to get into hot water with ati...

[perry_78]

Hmm, if this is legit, then we have to keep in mind that ATI has a tendency for less than polished drivers at launch

edit. though, if it isn't legit, which is the opinion I'm leaning towards, it's a pretty good money making idea, with the sheer amount of google ads

[Ubermann]

Anyone know when we get some DX10 benchmark programs ?

[DilTech]

The januari edition is a rippoff in my eyes.
Why buy that card for nearly 550$
When they offer a better card after only 2months omg.

$550?... Try $630 for the card in january.

Hmm, if this is legit, then we have to keep in mind that ATI has a tendency for less than polished drivers at launch

Bare in mind the 8800GTX driver is pretty much an infant itself, anything but polished, and definitely not anywhere close to it's top performance.

I'd say we have ourselves, for the first time in a long time, a REAL fight on our hands. IQ wise(if this review is real), performance wise, and feature wise both camps have a real beast on their hands.

On a side note though, people who already bought a 8800GTX need not feel buyers remorse.

[ownage]

If i look at the WHOIS output i have the idea the owner wants his privacy, probably because voiding his NDA contract or so.

[phi|os]

If i look at the WHOIS output i have the idea the owner wants his privacy, probably because voiding his NDA contract or so.

The owner better, or I fear for him.

[K404]

Stop complaining about the price

Its kinda reassuring that the graphs dont show a clean-sweep for the `600...if they`re willing to admit the 8800GTX wins some and comes close in some tests, adds come credability?

[Cybercat]

This article is bogus.

[perry_78]

This article is bogus.

strong words!

Either way, opinions are but speculation

[XSAlliN]

* DX10 full compatibility with draft DX10.1 vendor-specific cap removal (unified programming)

Now that's something new - DX10 whit Full support at release and DX10.1 latter, by 2008 will see even DX10.3 on this cards, I used to wonder why wasn't this possible before - guess now it's a thing of the past.

[Sanborn]

Yeah maybe if this came from a real website I would believe it......... Some 13 year old kid ion his basement probably made this. The writing in the article is terrible.