weberdarren97 :
I don't agree with this. You did the math correctly, but applied the answer to all AIB cards, and that's where I lose sight of the practicality of this equation. Because who's going to be using a RX 480 and GTX 1060 in the same system? I mean, I wouldn't get that combination even if Ashes of the Singularity was the only game I played.
No one suggested that you would be using a 1060 and 480 in the same system. The info is provided so one can decide
**which** of the two to put into your system.
Don't know how long you have been doing or following this .... but, if you have been following this for a few years:
1. the range of performance between AIB cards is not that great
2. the difference between cards has shrunk with each new generation due to nvidia limitations both legal and physical.
3. The advantage of the super cards (Classy, Lighting, AMP Extreme, Matrix, etc) no longer justifies their additional cost.
Since we don't have data for the 10xx series, let's look at 9xx
Gigabyte G1 131.4%
http://www.techpowerup.com/reviews/Gigabyte/GTX_980_Ti_G1_Gaming/33.html
Palit Jetstream 129.7%
http://www.techpowerup.com/reviews/Palit/GeForce_GTX_980_Ti_Super_JetStream/33.html
Asus Strix 128.4%
https://www.techpowerup.com/reviews/ASUS/GTX_980_Ti_STRIX_Gaming/33.html
MSI Gaming 127.2%
http://www.techpowerup.com/reviews/MSI/GTX_980_Ti_Gaming/33.html
Zotac Amp 127.1%
http://www.techpowerup.com/reviews/Zotac/GeForce_GTX_980_Ti_Amp_Edition/33.html
EVGA SC 123.6%
http://www.techpowerup.com/reviews/EVGA/GTX_980_Ti_SC_Plus/33.html
Reference 109.4%
https://www.techpowerup.com/reviews/NVIDIA/GeForce_GTX_980_Ti/34.html
So what we can we gather from the above....
-The average of all the AIB cards (excluding the EVGA) is 129% .... the variance is therefor just 2%. I will of course grant you therefore that the use of just a single AIB card is subject to an expected variance of + or - 2%
-The EVGA, with just a reference PCB and improved cooler, would not be expected to perform on par with the rest of the group.
-OTOH, we see a whopping 20% difference between the reference card and the group average. I think we can both agree that 2% pales in comparison to 20%.
Fury - X 105.05%
http://www.techpowerup.com/reviews/AMD/R9_Fury_X/34.html
Fury 107.75%
https://www.techpowerup.com/reviews/Sapphire/R9_Fury_Tri-X_OC/33.html
390x 107.12%
http://www.techpowerup.com/reviews/MSI/R9_390X_Gaming/33.html
390 108.21%
http://www.techpowerup.com/reviews/Powercolor/R9_390_PCS_Plus/33.html
Average = 107.03%
If we look at the AMD history, we see again that no card varies more than 1% from the average except for the FuryX which is of course a reference card
So again, I will grant you that, based upon past history, we have no reason to expect that any card will vary more than 2% from what we see here. Simply put, a) we have not seen an AMD card OC into double digits in a very long time and b) we have not seen a wide discrepancy between AIB cards from either nVidia or AMD and c) given what we have seen, there's no reason to expect that will change now.