Radeon HD 7000 Revealed: AMD to Mix GCN with VLIW4 & VLIW5 Architectures

[kaktus1907]

AMD Radeon HD 7000 fiasco as Bulldozer - likely Yes!
AMD sent to several manufacturers info-slides about the new Radeon HD Graphics 7000. According to first estimates results in the slides, AMD-fans will be disappointed. Game performance against the current ty-slow cards with Cayman not significantly increased, but consumption is relatively high for an advanced 28nm process. Stay tuned .... some slides will be here soon!

PS. Specs floating web in these days ARE all BS! ! ! ! ! ! ! ! !


obr hits again..

[Picao84]

AMD Radeon HD 7000 fiasco as Bulldozer - likely Yes!
AMD sent to several manufacturers info-slides about the new Radeon HD Graphics 7000. According to first estimates results in the slides, AMD-fans will be disappointed. Game performance against the current ty-slow cards with Cayman not significantly increased, but consumption is relatively high for an advanced 28nm process. Stay tuned .... some slides will be here soon!

PS. Specs floating web in these days ARE all BS! ! ! ! ! ! ! ! !



obr hits again..

It wouldnt surprise me. We have to keep in mind the change in architecture, with hardware schedulers, instead of a software one on VLIW. That has to account for higher sized die (compared to one in VLIW) as well as higher power consumption. However, if the upgrade in performance is little, I will also say they didnt went overboard with die size. They are following the carefull strategy of not doing big chips. To keep the die size low, they traded a big increase in performance for the scheduller, not risking doing both this generation.

[flopper]

It wouldnt surprise me. We have to keep in mind the change in architecture, with hardware schedulers, instead of a software one on VLIW. That has to account for higher sized die (compared to one in VLIW) as well as higher power consumption. However, if the upgrade in performance is little, I will also say they didnt went overboard with die size. They are following the carefull strategy of not doing big chips. To keep the die size low, they traded a big increase in performance for the scheduller, not risking doing both this generation.

it be the next great thing since christmas.
with the die shrink it will allow for greatness.
My faith is unwavered.
I am xtreme.

[Olivon]

Radeon HD 6990>GeForce GTX 590> Radeon HD 7970>GeForce GTX 580>Radeon HD 7950>GeForce GTX 570>Radeon HD 6970

http://www.donanimhaber.com/ekran-ka...-bilgileri.htm

[boxleitnerb]

http://www.donanimhaber.com/ekran-ka...-bilgileri.htm

Would be quite dissapointing if true. But also in tune with what Nvidia did with GT200 when they put stronger focus on GPGPU and the professional segment.

[labs23]

Would be quite dissapointing if true. But also in tune with what Nvidia did with GT200 when they put stronger focus on GPGPU and the professional segment.

If true...
Maybe a bit disappointing; should've been at least on par w/ 590, but IMO I won't be expecting something like this yet(prove me wrong here AMD).XD
Don't know much about GPUs BUT...as long as it'll drive 580's price significantly soon then I'll be one happy dood. LOL.

[flopper]

Would be quite dissapointing if true. But also in tune with what Nvidia did with GT200 when they put stronger focus on GPGPU and the professional segment.

without risking a big die the way to go, but 28nm should offer enough die to add for a nifty card.

[Mechanical Man]

Would be quite dissapointing if true. But also in tune with what Nvidia did with GT200 when they put stronger focus on GPGPU and the professional segment.

Why? Did anyone really expect over GTX590 theoretical performance? It even does not imply if its 1% difference on 590 or 580... Basically it does not say anything other than very coarse range where it will fall.

[fellix_bg]

It wouldnt surprise me. We have to keep in mind the change in architecture, with hardware schedulers, instead of a software one on VLIW. That has to account for higher sized die (compared to one in VLIW) as well as higher power consumption. However, if the upgrade in performance is little, I will also say they didnt went overboard with die size. They are following the carefull strategy of not doing big chips. To keep the die size low, they traded a big increase in performance for the scheduller, not risking doing both this generation.

CGN won't be much better than the legacy VLIW architecture in terms of efficiency. CGN is more like the current VLIW without the clauses, but keeps a very minimalistic approach to HW scheduling. It's still way behind what Nvidia have done ever since G80 in therms of scheduling and reordering, especially considering what Fermi can do now.

[LordEC911]

30cu/1920 sp units probably 7950, if it's not fake..

I thought CUs had to be disabled in quads?
No way it is 7970 since Tahiti XT CU count is not divisible by 3.
I was thinking 28CUs 1792SPs.

well, if we look at the driver, pitcairn & tahiti are supposed to be the same architecture. just as a hint.

Correct.

http://www.obr-hardware.com/2011/12/...rama-soon.html

and i can confirm that this is real.

Interesting...

CGN won't be much better than the legacy VLIW architecture in terms of efficiency. CGN is more like the current VLIW without the clauses, but keeps a very minimalistic approach to HW scheduling. It's still way behind what Nvidia have done ever since G80 in therms of scheduling and reordering, especially considering what Fermi can do now.

That's contrary to what I have been hearing, it sounds like a pretty big jump for AMD's GCN, just more moderate than what Nvidia has been doing.

[iMacmatician]

AMD Radeon HD 7000 fiasco as Bulldozer - likely Yes!
AMD sent to several manufacturers info-slides about the new Radeon HD Graphics 7000. According to first estimates results in the slides, AMD-fans will be disappointed. Game performance against the current ty-slow cards with Cayman not significantly increased, but consumption is relatively high for an advanced 28nm process. Stay tuned .... some slides will be here soon!

The key here (given the quote posted here and if this report is true) is the "consumption is relatively high" part. Since that statement is talking in relative terms, there's nothing to base the "Game performance … not significantly increased" statement. The chip they refer to could be a "7800" shrunk Cayman and the "7900" top-end GCN could be a lot faster than Cayman. But, even if they were referring to the "7900," I wouldn't consider it that unexpected.