Fermi

[Mush Brain]

The fastest architecture technically?

Allow me to explain.

A GTX 580 at 1ghz beats a 660Ti at the same clock with far fewer cores "580"

I know that they are different architectures but how is Fermi this fast at a lower clock and lower core count?

I am beginning to think that Fermi under 28nm with more cores would have really took the piss vs GCN, as Fermi was a compute powerhouse vs VLIW.

Efficiency would possibly be the same as 40nm Fermi due to added cores and a higher core clock.

Any input on this idea?

[Sonefiler]

no

[TheMidnightNarwhal]

yes

[werto165]

maybe

[H4X3R]

idk

[Guest]

WTF?? The responses so far.  

 

If only the first and second post were switched.