NVIDIA GeForce 7900GT Pics, Specs

[meson2000]

I am still kinda wondering what the point of HDMI on a PC is. It must only be for home theater PCs that you intend to hook up to your home theater. Otherwise it makes no sense. DVI-HDCP has to be the standard for regular PCs to watch Hi-Def content.

Why on earth would I need HDMI on a regular PC? So I can run my audio into my monitor and be forced to watch my Hi Def content and listen to the sound through the crappy integrated speakers in my monitor? Yeah.. that makes sense....

 

[coldpower27]

Originally posted by: Regs
It was a loose reference in stating the fact that they only work with what they can get.

Everyone works with what they can get.. this doesn't tell me anything new. :roll:

 

[beggerking]

Originally posted by: coldpower27

Originally posted by: Acanthus
Ok, so in short you determined a die shrink can make the die smaller :roll:

What I determined is that if it is indeed a simple die shrink the cost advantage will be in Nvidia's court for sure as 224mm2 will yield much better then the R580 cores 315mm2.

Nice calculations..
just wondering... how is die size affecting yield? is there a direct relationship between them?

 

[beggerking]

Originally posted by: coldpower27

Originally posted by: Avalon
Acanthus, if Steelski's math is right and going from 110nm to 90nm is a 34% decrease, then I don't see how adding 8 pipes would be feasible, as it would take up the majority of the free space earned from the die shrink. Plus, other things they need to add to the card. I'm not sure, I could be wrong. Thoughts?

Well you got to think about it how much would adding 8 Pipelines cost....

Adding 8 Pipelines from NV40 to G70 was 80 Million Transistors, but the pipelines were also enhanced as well, also 2 more vertex units were added, enahnced buffer for 20x15 operation, if we be generous and give that amount for 8 pipeline that would bring the total to 382 which ~ the same level as R580 so it is possible to produce for Nvidia, if ATI can rpoduce a 384 Million Transistor GPU on 90nm why can't NV do the same?

agreed! so should we be expecting a 32pipe gtx?

 

[coldpower27]

Originally posted by: beggerking

Originally posted by: coldpower27

Originally posted by: Acanthus
Ok, so in short you determined a die shrink can make the die smaller :roll:

What I determined is that if it is indeed a simple die shrink the cost advantage will be in Nvidia's court for sure as 224mm2 will yield much better then the R580 cores 315mm2.

Nice calculations..
just wondering... how is die size affecting yield? is there a direct relationship between them?

The larger the die the more chance there is that an defect will occur on the die because of the simple fact you have a larger surface that has to be error free, well let's put it this way, also with a smaller die you can also get more chips per single wafer, as the wafer size is constant, even if the same amount of chip were defective, you will still yield more.

For instance on 300mm Wafters you can get optimally 179 Chips of G70 Core, or 274 Chips of G71 Core assuming it's an optical shrink of course, this is at 100% Yield Rates which don't occur, you always have some defect so yields are going to be lower. Having a smaller core though helps reduce costs, as you can get more of them per wafer, if % yield rates are the same of course.

Also, since the projected specs of G71 are 650MHZ/1.6GHZ, this allows Nvidia to used cheaper GDDR3 1.2ns which is much more available then the 1.1ns chips have been.

 

[coldpower27]

Originally posted by: beggerking
agreed! so should we be expecting a 32pipe gtx?

I am not sure that we should be expecting one, as you be setting yourself up for disappointment if you did that, IMHO, I would think that Nvidia will take this round to be more cost friendly and produce cheaper cards with good performance but not necessarily the overall crown, and focus their efforts more on G80 and making sure that is an extremely good product.

32 Pipes is doable on 90nm, just I don't expect Nvidia to endure another over 300mm2 die cycle, the R580 is ATI's first product greater then 300mm2 so they are ok for this round.

I would like to be presently surprised though and am hopeful the G71 is a 32Pipe monster, as if clocked at 650MHZ that would one fast card and not just in new games, but old ones as well. At 24 they will simple be offering ~ the same level of performance as ATI but at cheaper cost to themselves, which is also a good thing for THEM and perhaps for us if it translate to lower prices and wide availablity.

 

[gobucks]

I have to say that I think ATI has won this round in terms of performance, but nvidia is still gonna get my money cause the price is right. I can't believe there is gonna be a 24-pipe, 450MHz card for $299 MSRP!!!! I think this might be a good time to upgrade my 6600GT, lol. The GTX is looking a bit disappointing, cause it's a pretty sure thing that there will only be 24-pipes, not 32. after all, if it did have 32 pipes, it'd have around twice the pixel fill rate of a GT, making it a better value than the GT. The highest end GPU is NEVER the best value from a price/performance point of view. But i think its price will attract a lot of people.