nice looks like my next buy
Printable View
nice looks like my next buy
http://tof.canardpc.com/view/c9c19e7...aa08cc250e.jpg
Well, this card looks to be the same size board as GB 460SCO version, the rumor of board partners able to use similar PCB's appears to be true.:)
The card looks identical to the 460 @ newegg now: http://www.newegg.com/Product/ImageG...20Video%20Card
Now I'm almost sure we will get some short GTX 560's for SFF builds. Can't wait for lanch.:D
http://img141.imageshack.us/img141/6...65x5osf95k.jpg
http://bbs.expreview.com/thread-39288-1-1.html
Quote:
Voltage to 1.15v, modify the BIOS does not increase the higher voltage. Up to 1065 has been a core. Oh very good. AMD at this voltage it can not run this frequency. Wow haha.
Is valid and cards, you know, but absolutely can not say oh. Haha.
woot 1065 MHz @1150mV!! my 460 only can run 925MHz @1150mV
%20 faster than my GTX 460 945/4680MHz @ Graphics score
http://3dmark.com/3dm11/72044
%6 lower than HD6950 @925/5800MHz @Graphics score
http://img231.imageshack.us/img231/4...k11details.jpg
Like I expected, these things will be amazing overclockers.
HD6950@6970 with same good cpu
http://i82.photobucket.com/albums/j2...5P6218info.jpg
P6218 and 1075/1475
+17.7% higher ;)
Yeah, 560 is the 3th in nVidias single-GPU line up (560> 570>580 line), and actually falls into mainstream, but AMD has chosen to use it's 2th GPU 6950 and "degrade" it to a 1GB-mainstream to fight it.
That's not a fare fight, LOL, and then AMD may even get the upper hand in performance (maybe, but we have to wait and see).
But even then, the mainstream-battle is not about performance-crown, it's all about price/performance.
The battle is going to be fought in price-arena, in this round.
If I'm not mistaking, nVidia is going to price this well, but nVidia shouldn't have a big problem to put a big blow on 275$, after all this their mainstream card. We have to wait and see how aggressive they will get, tho.
I don't care about die size, yields, number of transistors, etc .. I don't know if these factors affect the manufacturing costs either.
As far as I know, they use most on R&D, not manufacturing, but anyways why should I care? Shareholders and those on payroll should care about these stuff.
As a consumer, I don't know, and don't care how much it cost to produce these cards. I only care about the price/performance/power usage for a mainstream card.
ok so the best of the GF114 vs the worse of the Caymen and both are about the same size
and that means one should have much more room for a price drop thats still profitable.
Profitable for who? Are you a shareholder or on the payroll? If not, why do you care?
AMD has started with 275$, just $20 under the price they were asking for high-end card. (It could actually be unlocked to the highest end)
Are you saying that nVidia vil have problems to put a big blow on it with a mainstream card?
We can speculate on number of transistors, dies size, yield, etc .. but you need to see the different between high-end and mainstream pricing to catch this point.
true,
so perf/watt is important since it means less: heatsink, vrms, PSU, and electric bill.
ive done the math before and comparing the mid range, the difference between good and bad (in real cases) came to 25$ across 18 months, which is quite significant and definitely a factor in purchasing.
between more common things the difference is about 5-10$ for the one with a little worse perf/watt, and thats just to the electric bill.
and the memory is probably not that much different, and not going to impact profitability nearly as much as the chip size does.
in the end the architecture is going to decide so many of the little things that end up going into the final cost.
looking at just price and just performance can give you what is now or what is predicted for a launch, but offers very little forecast into how it will evolve in the upcoming months when products shift, eol, or drop in price.
Cayman is 2.64 billion transistors and gf104 is 1.95 billion. Why is it that their size is so similar. As far as transistors/performance NV is better. But they lose what matters, the die size per transistor battle.
Why is fermi so inefficient in this regard?
I've seen this kind of component-cost arguments before.
It's a tricky one, because only GPU-designer directly involved in development can talk about it.
I don't think we should speculate on stuff nobody has a deep and correct understanding about.
Each of us can speculate and would never agree, because only a few knows what's really going on in nVidia or AMD's R&D or fabs.
So lets cut these speculations and talk about how many $ we have to pay for the performance. That's what we all understand, and that's only thing a consumer cares about, actually.
Wow sweet freaking card, its looking like nvidia has amd across the entire range this should trigger the price cut card :P
Nvidia would have to almost equal 6970 with 560 to be on the same level if you consider die size. GF104 (and GF114 presumably) is only slightly smaller than Cayman so cost per chip is about the same from a manufacturing point of view. In that sense Nvidia is still behind AMD, but GF114 will certainly bring them closer than what they were with GF104. And being close means more competition -> lower prices. :clap:
To be honest, I don't know, and I don't care.
I don't think you know either, maybe you have heard from somebody somewhere. Then you repleted it until you believed it yourself.
Are you an engineer in nVidia'a R&D? Or are you directly involved in nVidias fabs? If not, how do you know? these guys keeps these stuff as a business secret.
Nowhere, nowhere you can see a official document that nVidia or AMD explain these stuff.
tell me how do got all these detail info about nVidia's biasness secret?