Are benchmarks becoming more biased against AMD?

dude807

Distinguished
Jun 19, 2013
215
0
18,680
Hey guys. I've looked at probably hundreds of benchmarks, some on toms hardware, some on linustechtips and from various websites. One thing that i've noticed is the r9 290 and r9 290x always seem to be left in the dust with the new GTX 980's, even when they're both overclocked. I tested this myself and i found that my r9 290x performs better than the gtx 980. Even its Furmark score is better. I averaged 75 fps and the 980 averaged 72. So are these benchmarks biased, or am i just getting lucky? I'd love to hear what you guys think. :]
 
Solution
I think part of it too is the drivers, not sure if the sites update the drivers on cards they've tested in the past, and that can make a huge difference. The 290X has been out for a while, I love mine, but in the beginning the drivers weren't all that great, there have been a LOT of improvements since then, the 980 is new, so the drivers are still babies. The performance of the 980 will only get better from here out, while the performance now of the 290X is likely to be peaked. 980 with crappy driver VS. 290X with good drivers, it's not a surprise the 290X performs better on some tests.

iamlegend

Admirable
In terms of Raw performance and real gaming scenarios, I think there is no such a big gap between those 2 cards, the main thing is that the 980`s consumes less power hence delivering a superb performance.

However, I have red a thread/review which compares those 2 cards and the findings is that when it comes to squeezing the performance from 980 it does not break the law of physics (power-performance) which indicates that the claim of having less power to more performance is inaccurate.
 

dude807

Distinguished
Jun 19, 2013
215
0
18,680


Most benchmarking builds i've seen are much better than mine, with high end Cpu's and motherboards, yet i still beat their results or come within a few fps which could have been caused by my lack of cpu power. The furmark test is the lease bias though. It depends souly on the gpu. So you can see why i'm a little confused and suspicious.
 
I think part of it too is the drivers, not sure if the sites update the drivers on cards they've tested in the past, and that can make a huge difference. The 290X has been out for a while, I love mine, but in the beginning the drivers weren't all that great, there have been a LOT of improvements since then, the 980 is new, so the drivers are still babies. The performance of the 980 will only get better from here out, while the performance now of the 290X is likely to be peaked. 980 with crappy driver VS. 290X with good drivers, it's not a surprise the 290X performs better on some tests.
 
Solution