Originally posted by: Azn
Originally posted by: MODEL3
I'm not even going to bother replying you again!
Because for all those "points" that you have in your latest reply, I have answered analytically already!
It is pretty clear that we cannot find a common ground, so let's just say that, we agree to disagree!
Anyone who is reading this topic, can see the points each of us made, so he can make his own conclusions!
Everyone here aren't dumb. At least the forum frequents who's been around. you don't have to repeat entire history of GPU. There was no point. You just changed the subject and created more subjects for you to make a point of something we weren't even discussing. On top it all off you debunked your own supposedly branched out points.
I asked you simple question why you think the chip is big but what I got is rhetoric and strawman. Then say the chip isn't big. It's pointless. You say things without meaning what you say.
Is your effort to put words in my mouth, ever going to stop?
This is your practise from the start!
It is clear to anyone, that i said that i have faith that the people who are reading this topic,
can understand perfectly, so they will make their own conclusions!
Stop changing the meaning of what i say.
You are arguing about things that nobody implied!
Can't you understand that everybody knows your practice by now?
---------------------------------------------------------------------------
Whenever I said something logical you replied with like:
What 1+1=2?
What? What is this? More rhetoric? For crying out loud!
---------------------------------------------------------------------------
I don't even know where to start:
Let's take your scenario that the site's 5870 is 300mm2 (I said 350mm2 with decoupling transistors)
You said that:
if NV can launch 8800GT and make a profit with the 8800GT (334mm2) at 250$
why can't a 5870 (300mm2) have 300$ price at launch?
Originally posted by: Azn
So you agree a 8800gt that has the 334mm die size sold for less than $300 and still making good profit. Here you are trying to justify yourself by differentiating high end or lower end when the die size is the cost of the chip and whatever left is profit
-------------------------------------------------------------------------------
You comparing the launch price of a 8800GT model,
with what launch price should have a 5870 model, in order to make your points!
-----------------------------------------------------------------------------------
For the G92 chip (334mm2) we had also the following higher models (9800GTX, 9800GTS)
So we had:
9800GTX (the highest G92 model with Two 6-pin power connectors)
8800GTS (a lower clock G92 model within the 150W power range)
8800GT (a even lower clock model with some TU&SP disabled)
What you should have done, is to compare the price of 8800GT,
with the price of
the analogue ATI model which is the 5830!
For the RV770 (256mm2) we had:
HD4870 (the highest RV770 model with Two 6-pin power connectors)
HD4850 (a lower clock RV770 model within the 150W power range)
HD4830 (a even lower clock model with some TU&SP disabled)
Let's make the analogue product line for the RV870 (or whatever is its real name)
HD5870 (the highest RV870 model with Two 6-pin power connectors)
HD5850 (a lower clock RV870 model within the 150W power range)
HD5830 (a even lower clock model with some TU&SP disabled)
Can you tell me, if you can comprehend now that the
parallel model for 8800GT is not the 5870, but the 5830?
Or it difficult for you to understand this also!
-----------------------------------------------------------------------------------
Now let's see
the launch pricing:
HD4870 300$
HD4850 200$
HD4830 160-150$ (actually it was 130-120$ but it launched later than 4870/4850)
Lets see the analogue launch pricing for the 5800 series:
I will take my scenario, which I said that the die of the site's 5870 would be something like 350mm2,
and according with the pricing history of ATI's should be priced 400$ and lets scale the prices:
HD5870 400$
HD5850 270$
HD5830 210-200$
Even with my high priced scenario, the 5830 is around 50$ less than the launch price of 8800GT(250$)
So for the example you gave (8800GT),
ATI's profit per die size is cheaper than NV's
If I take your scenario that the die is 300mm2 and that the launch price of the 5870 can be achieved at 300$:
HD5870 300$
HD5850 200$
HD5830 160-150$
So with your scenario, you can see clearly, that the 5830 is around 100$ less than the launch price of 8800GT(250$)
So again, for the example you gave (8800GT),
ATI's profit per die size is
extremely cheaper than NV's!
Even if i take a completely unrealistic scaling like the below:
HD5870 300$
HD5850 250$
HD5830 200$
it doesn't change the fact, that the 5830 is around
50$ less than the launch price of 8800GT(250$)
And again, for the example you gave (8800GT),
even with this completely unrealistic scaling,
ATI's profit per die size is
cheaper than NV's!
------------------------------------------------------------------------------------
So from the start you where trying to suggest:
why can't the 5870 be in the price level or a bit higher than the launch price of a 8800GT?
And I proved you, that with a
300$ 5870 launch price,
the equivalent product (5830)
is way cheaper than 8800GT!
Even for my own scenario (350mm2),
I proved you, that even with a
400$ 5870 launch price,
the equivalent product (5830)
is cheaper than 8800GT!
And you think that, I have comprehension problems?
Originally posted by: Azn
Do you have comprehension problems?
-----------------------------------------------------------------------------------
Do i have to go to this length, for every point you made, to prove that your points are not valid?
------------------------------------------------------------------------------------
You don't even seem to understand that the profit of the G92 based cards,
is not the profit of the 8800GT model,
but it is the
weighted average of the profits of:
9800GTX/8800GTS/8800GT&8800GS models!
-------------------------------------------------------------------------------------
Also, have you checked the Financial state of Nvidia lately?
------------------------------------------------------------------------------------
Originally posted by: Azn
Originally posted by: MODEL3
The GT260 launched at 400$ and the GT280 launched at 650$.
So you have to calculate the weighted average for the selling price of GT200 series / per quantity / per model!
So it was not 300$ at all!
In fact it was more close to 500$
Launched for how long? 2weeks? That's when 4870 and Nvidia matched their price by cutting GTX 260 and GTX280 price by $100. Even for those who bought it at launch price people got refunds.
In your last reply, you said the above!
I will repeat the question I asked you in my early, early replies:
Nvidia lowered the original launch prices of GT200, because of 4870 (like you said above),
so essentially Nvidia had
lower profit per die size with these new prices!
What is the reason now, that forces ATI, to design a GPU
with lower profit per die size in relation,
with the profit that they had in the past 2 years?
Also if the rumours are true,
isn't going ATI to have the first DX11 solution in the market?
Why they need to lower their margin, when they can make more money based on this fact!
Don't get me wrong it would be nice, for ATI to lower their margin even more!
But this is not about, what i or you want.
We are talking about what ATI wants to do!