Quote:
Originally Posted by SonDa5 
I thought they were cherry picked for low ASIC based on the ASIC score rating.
My MSI TF3 PE/OC GTX 560 ti 448 was scored in the 70s on the ASIC scoring yet could clock to 1GHZ on air with stock voltage. I don't put alot of trust in the accuracy of the ASIC score program used in GPU-Z. DOesn't really seem to be accurate of real gpu performance.

I thought they were cherry picked for low ASIC based on the ASIC score rating.
My MSI TF3 PE/OC GTX 560 ti 448 was scored in the 70s on the ASIC scoring yet could clock to 1GHZ on air with stock voltage. I don't put alot of trust in the accuracy of the ASIC score program used in GPU-Z. DOesn't really seem to be accurate of real gpu performance.
Well as has been said before, asic quality is only one factor. In no way should it be taken as an determination of how well your chip will overclock. It is still all about the silicon lottery. Gpu-z correctly reports asic % as it is programed into the chip itself once the leakage is determined at the foundry (it is how programs and ccc itself can know what voltage to send to the chip).
What asic % helps us with is knowing how much voltage (thus heat and power) a chip will produce. For air users, a higher asic can mean lower heat and thus possibly a higher overclock than a lower asic % with higher voltage. As you said, you had low asic and high clocks, so it is not always the case.
Cherry picking gpus doesn't make much sense only because it is difficult to test every chip by pushing it to its limits. By merely setting a clock level at stock volts,you can effectively skim more chips quicker and with much more consistency. AIBs don't like changing voltages if they don't have to, it just becomes messy.






















