Carzy Zarx’s PC Gaming Emporium - Catch Up on All the Latest PC Gaming Related News

[JEMC]

fatslob-:O said:

JEMC said:

I let it go in your previous post, but you're doing it again: don't try to cheat comparing reference vs overclocked cards!

Sure, a Titan X overclocked to 1.2GHz can be 60% faster than a reference 980. But a reference 980 is clocked at 1.1GHz and can be "easily" overclocked to almost 1.4GHz, with the boost exceeding 1.5GHz, reducing that advantage to what, 35-40%?

That's the performance jump Nvidia delivered going from the 780 to the 980, with two different architectures but the same process node and going from a 560mm² chip to a smaller 400mm² one.

With a new architecture and a new process node, Nvidia can deliver that performance jump with a 300mm² chip.

Although Nvidia would like to market that improvement very much!  

I don't know if it's totally fair to use an extremely mutilated GK110 part since the 780 Ti/Titan Black was much much closer to the GTX 980 ... 

If you want, we can compare the jump from the GTX 580, a refined 520mm² Fermi chip using 40nm to the GTX 680, a less than 300 mm² cut down Kepler card build using the 28nm process. The jump in performance was also between 35-40%.

That would be the ideal comparison as there's an architectural change as well as a process node improvement.