NVIDIA Reports Financial Results for Second Quarter Fiscal 2011

[AndreiD]

Then my bad and sorry for the sarcasm.
Xbitlabs and the others took the TechReport number for granted.
And if you read the TechReport article carefully, they say that they estimate that the die size is that big.
I believe the original chinese article where they machined the GTX460 heatspreader off and actually measured the die size and they got 366mm^2 (NordicHardware didn't make the measurements, they just reposted what a Chinese website found).
Fudzilla is like a nvidia heaven and Semiaccurate might seem sometimes against NV but what they reported so far isn't far from the truth, in the end, you read what you want.
Anyway, it's not like the die size matters in any way for us the consumers, it just matters for the bottom line of the chip companies.

 

[Benetanegia]

Semiaccurate might seem sometimes against NV but what they reported so far isn't far from the truth, in the end, you read what you want.

Hmm they are always against Nvidia and sometimes (i.e a few times) are right. My concern with them is that they do not do anything close to journalism, yet they pretend to be a news site. A news site has to be always right not just 50% of the times (how do you do it? easy, you actually have sources). It's easy to be right 50% of the time if you write like they do. i.e. "GF100 is going to be big (everyoe knows that), so it will consume a lot and be hot". Both things go hand in hand, so it either is hot or it isn't. They were right that time, BUT and it's a big "but" there, they said the same abut G80, GT200 and recently GF104. None of them was what they claimed to be. 25% accuracy right there. If you are not right 90%++ of the times and you are right just a little bit over 50% of the times (when you have a good day) what are you? Nothing, since throwing a coin is just as accurate as you are.

Anyway, it's not like the die size matters in any way for us the consumers, it just matters for the bottom line of the chip companies.

Yeah, but it does not matter nearly as much as many people claim it to be. And frankly most of those people just repeat what Charlie says in his editorials full of FUD and BS, which usually put GPU manufacturing costs at above $300 which is a total load of BS. In general, for high end cards the difference between a small chip and a big one is $25 at most, not a problem at all, specially when the bigger chip is faster and hence can be sold for more. That's without taking into account many other expenses, which in the case at hand Nvidia deals better with, like management. AMD is doing better right now, but it was a mess just a few year ago and isstill not at the level of most other IT companies on that department. For example, had Nvidia not have to pay 190 millions for the expensees derived from bad bumps in 2008, they would have been $50 millions in the black, more than AMD's graphics division (who apparently sells HD5000 cards like hot-cakes and a massive profits margin, but doesnt show on the finances). And that's without having the products to sell, except GF100 based cards which according to the same sources that are claiming GF104 to be 367 mm^2, are selling at a big loss. Erm can you tell me how can a company sell 400.000 cards at a loss and have a 50$ profit in a bad quarter?