AMD: Southern Islands (7*** series) Speculation/ Rumour Thread

[Love_In_Rio]

So 2 full Tahitis @ 850 mhz with only 300W TDP?
(and probably a bios switch making it a full 7970x2 around 375W)
That sounds pretty efficient

Or two full 7980s ( Tahiti at 1/16 rate DP )

 

[denev2004]

Well, thats only 3GB per chip, much like all their other cross fire on a stick cards.

Given that theres already a 6GB for ONE GPU card coming out, 6GB for two GPUs is hardly amazing.

Exactly. It's just a normal improvement over the previous cards, there's nothing special.
But I wonder whether 850Mhz is a bit low.

 

[tunafish]

But I wonder whether 850Mhz is a bit low.

Probably necessary for 300W. They are free to provide a buff enough VRM that ocing it to 1GHz should be a breeze.

 

[Acert93]

Where is the 300W figure coming from?

 

[TKK]

Where is the 300W figure coming from?

I wonder the same. I doubt we'll see anything less than a 375W TDP.

 

[AnarchX]

Maybe AMD will reconfigure PowerTune as Boost:
- 925MHz dual GPU boost, when front-end is limiting
- 1100MHz single GPU boost, in cases when AFR is not working, through render-to-texture or similar

Of couse enthusiasts should be able unlock it to >375W and a included water-cooler would be appropriate if they are aiming >$800.

 

[Silent_Buddha]

If they have the power circuitry on board for this, then it should theoretically be capable of overclocking to the same limits as a 7970.

The clocks are probably just there to keep TDP under a certain level for certification.

Regards,
SB

 

[Zaphod]

Where is the 300W figure coming from?

I wonder the same. I doubt we'll see anything less than a 375W TDP.

Maximum PCIe power spec. If they want the card to be certified, they won't go above 300W @ stock settings.

 

[cal_guy]

Maximum PCIe power spec. If they want the card to be certified, they won't go above 300W @ stock settings.

That's not really true

At the end of the day as the PCI-SIG is a pro-compliance organization as opposed to being a standard-enforcement organization, there’s little to lose for AMD or their partners by not being compliant with the PCIe power specifications. By not having passed compliance testing the only “penalty” for AMD is that they cannot claim the 6990 is PCIe compliant; funny enough they can even use the PCIe logo (we’ve already seen a Sapphire 6990 box with it). So does PCIe compliance matter? For mainstream products PCIe compliance matters for the purposes of getting OEM sales; for everything else including niche products like the 6990, PCIe compliance does not matter.

http://www.anandtech.com/show/4209/amds-radeon-hd-6990-the-new-single-card-king/18

 

[Tchock]

Hmm... chances of pushing a 1+Ghz 7950 1.5GB out there at $380-400, post-GTX680?

Now's a good time to do a balanced price-perf SKU that doesn't necessarily hurt AMD (1Ghz vs full-unit chips, I've no idea, but it'd probably be a good idea), while maintaining almost the same performance gap wrt the 680 (probably 15% or so, at a 25-30% price gap).



(I want my quarterrate dp fix )

 

[Zaphod]

That's not really true

AMD may or may not care about PCIe compliance for a dual GPU halo card, but that's almost certainly where the assumption (that may or may not be warranted) of max 300W comes from (which was the question).

 

[AlphaWolf]

Considering the range on power use for the 7970, I don't find it that hard to believe they could manage 300W, they can use lower voltage and aggressive binning. I also doubt it matters to most of the people that would buy one.

 

[Alexko]

From Tridam (Hardware.fr):

"AMD nous a indiqué hier avoir l'intention de proposer une version GHz Edition de la Radeon HD 7970, sans préciser la forme exacte que cette déclinaison prendrait."

Which you could translate as:

"Last night, AMD told us that they intend to launch a GHz Edition of the Radeon HD 7970, without specifying exactly what it would be."

I sure hope it's running above 1 GHz, otherwise the 8% overclock will leave most people underwhelmed.

 

[DarthShader]

Could be enough, if they can add a Boost on top of that. This could be doable with current hardware, but with new Bios. (or not?)

 

[Dave Baumann]

Could be enough, if they can add a Boost on top of that. This could be doable with current hardware, but with new Bios. (or not?)

Could be, but it is a question for each particular board target if it is necessary.

 

[hurleybird]

From Tridam (Hardware.fr):

"AMD nous a indiqué hier avoir l'intention de proposer une version GHz Edition de la Radeon HD 7970, sans préciser la forme exacte que cette déclinaison prendrait."

Which you could translate as:

"Last night, AMD told us that they intend to launch a GHz Edition of the Radeon HD 7970, without specifying exactly what it would be."

I sure hope it's running above 1 GHz, otherwise the 8% overclock will leave most people underwhelmed.

AMD could have had something like this ready for the Kepler launch.

With any luck we'll get an improved cooling solution, improved components, and ~1.2GHz stock clock.

Or maybe they'll try to lower voltage to improve perf/watt vs. Kepler.

Anyway, got a link?

 

[Alexko]

AMD could have had something like this ready for the Kepler launch.

With any luck we'll get an improved cooling solution, improved components, and ~1.2GHz stock clock.

Or maybe they'll try to lower voltage to improve perf/watt vs. Kepler.

Anyway, got a link?

To Damien's statement? It's in the comments section of this review, page 5, 3rd comment from the top: http://www.hardware.fr/articles/857-1/nvidia-geforce-gtx-680-test.html

 

[Deleted member 13524]

So.. erm.. any speculation as to when will AMD lower the MSRP of the HD7900s?

 

[Broken Hope]

Hope it's soon, I want to upgrade my 5870 but I'm not paying £400+ lower it to £330-360 and you'll have an instant purchase from me.

 

[Acert93]

Maybe someone can answer these questions for me. Is Tahiti unbalanced? Based on Pitcairn it seems to be less efficient. Does Tahiti have too much memory? Seems the extra 1GB isn't offering a performance or visual benefit. Too much bandwidth? For an extra 37% bandwidth it doesn't show up in benchmarks. Is the wide bus/memory controller eating unnecessarily into GCN TDP? Is GCN spending more on compute than Kepler or is Kepler inching out performance other ways per mm^2? How much can be chalked up to Kepler being launched about 3 months later and being able to tune their final product to Tahiti?

I am asking these questions after looking at the two chip "costs" and putting aside their architecture. I am more curious if it is an architectural issue or an implementation difference. GCN Tahiti 7970 is a bit bigger than Kepler GK104 (~ 352mm^2 / 4.3B transistors versus ~ 294mm^2 3.5B transisters) and, specifically, has more memory on a wide bus (384bit 264GB/s versus 256bit 192GB/s). The extra 1GB of VRAM may not be coming into play just yet with current software but is an advantage. Kepler has a slight edge as it is clocked higher at 1006MHz plus Boost verus 925MHz for GCN. So the general picture we see is that in theory GCN should be faster than Kepler:

Die Size: 19% larger
Memory Bandwidth: 37.5%
Memory Size: 50%
Frequency: (9%)

So my question is: What is holding GCN, specifically Tahiti, back? As others have noted it seems Pitcairn at 212mm^2 scaled up to Tahiti size would, in theory, be faster. So is Tahiti not "balanced" correctly?

And is the extra 1GB of memory and 384bit bus overkill? Are there any tests show that GCN is a more bandwidth needy design than Kepler? My understanding is that memory buses and the traces to the memory and extra chips costs a lot of power, plus they take up a lot of area real estate.

Can any of this also be chalked up to NV scaling back compute in their consumer graphic cards (an assumption I am making, right or wrong, based on how Kepler seems to fall far behind the Fermi cards in a lot of compute tests) and being able to fine tune their launch in response to GCN?

Edit: Maybe I should have made this a new thread? GCN and Kepler compared, contrasted or something as such?