NVIDIA GF100 & Friends speculation

Arty

KEPLER
Veteran
Here is the best picture of the Fermi board I can find, courtesy Tim Smalley. Thanks Tim :love:

8 pin + 6 pin = < 300W TDP

2z6yrzn.jpg
 
So what have we learned so far?

GF100 will be faster than HD58x0s.

GF100 will be bigger than HD58x0s.

It will be out by Christmas.

Charlie has a bigger ego than I thought.

Charlie is wrong again.

Anand has an unnatural love for Intel.


Question.....


Since when did the size of a chip mean anything?

If it is just as cool and uses just as much power and fits inside the same size cooling device as the comp..... who cares?
 
GF100 will be faster than HD58x0s.
By the time it launches? don't hold your hat.
GF100 will be bigger than HD58x0s.
bigger than the 500mm2 noted above
It will be out by Christmas.
Not quite
Charlie has a bigger ego than I thought.
Granted
Charlie is wrong again.
Not yet
Question.....

Since when did the size of a chip mean anything?

If it is just as cool and uses just as much power and fits inside the same size cooling device as the comp..... who cares?

You as an end-user should notice it as it's die size is an indication of the cost of the processor.

What I'm worrying about is how much will be done in software? Tesselation is kind of confirmed, but AA maybe? AF? And how will it handle older game code, i.e. non DX11 code? This is so much better than GT200, it could hurt from pre GT300 optimizations.
 
I'm pretty sure it can handle physics code damn well ;), so if we noticed in the past how chips that handle physics code well, how they handle game code hmm well that should be a good indication.
 
I wonder if AMD's stance will change soon from "Open Physics Standard is Better" to "No Physics on the GPU is Better"...

GPU is A1 35th week of 2009 (August?) by the way.


It probably will, AMD does have bullet though aslong as games start coming out with that, I'm sure AMD will jump on it.

Yep when was that tape out again, Chuck :D
 
Back
Top