Quote:
Originally Posted by brettjv 
Why does it matter to you so much if they aren't running at the same speed? Some very simple math will allow you to run them at the same clocks using Afterburner if you want to.
Not only that but all that's going to happen from the clocks being mismatched is that they'll either run at the average speed between the two of them, or they'll run like they would at the lower of the two clocks. It really shouldn't hurt anything at all.
It's also pretty obvious based on the wide range of Max OC's we're seeing that these chips vary by quite a bit. If they didn't have means to set the variable boost amount differently for different chip samples, then the boost clock on every one of these cards would have to have been set really low, i.e. based on least common denominator.

Why does it matter to you so much if they aren't running at the same speed? Some very simple math will allow you to run them at the same clocks using Afterburner if you want to.
Not only that but all that's going to happen from the clocks being mismatched is that they'll either run at the average speed between the two of them, or they'll run like they would at the lower of the two clocks. It really shouldn't hurt anything at all.
It's also pretty obvious based on the wide range of Max OC's we're seeing that these chips vary by quite a bit. If they didn't have means to set the variable boost amount differently for different chip samples, then the boost clock on every one of these cards would have to have been set really low, i.e. based on least common denominator.
I think you've nailed exactly why Nvidia decided to use a boost clock for Kepler. There is too much variance between GPU's this generation so they had to customize each cards clock speed to suit its own personal chip. Certainly allowed them to get away with clockspeeds they would never have defaulted to if all the cards were set at a static reference clock....























