The test configuration have the following elements:

Graphic drivers:
Before we review the results, we have to clear the way of tests. We did not poke the GeForce drivers’ and Catalyst’s settings at the synthetic gauges (3DMark Vantage, DirectCompute Benchmark, Unigine Heaven Benchmark) everything ran in factory configuration. After the synthetic tests we turned on manual 16× anisotrophic filter in both drivers.
Video cards, power usage, warming:
We gauged all participants for power usage and warming. We analysed the appetite of cards in Windows rest by watt metre, in oppress by FurMark 1.8.0 inside gauge module. We strived to constitute approximate average from datas at both time. Temperature maximum was observed by FurMark too.

Muscle dudes
Participants:
We are not concerned with the looks because we have studied them extensively in the past (articles or HOC TV broadcasts). Furthermore the Radeons are all reference models, only the GeForce is exception, that is an ASUS Matrix model with uniqe refrigeration and a taint signing lighting Matrix label. We would like to start in the game a GeForce GTX 295, but our largest grief it could not give a picture for our repeated attempts neither, so we sent it to the locker room.
ASUS EAH5850


ASUS EAH5870


ASUS EAH5970


ASUS Matrix GTX285


Received power usage values:
In case of cards like this calibre – especially talked by CrossFireX system – the power usage is secondary. Although these cards can throw up the electricity bill supremely, confess that if somebody buy one of these cards, he will not concerned by the bill. Good to see that the cards in rest eat slight, it stopped at 164 W by the HD 5970. In opress HD 5870 and GTX 285 is in same level, but HD 5850 eats less. The HD 5970 was the most hungry in the competition, hungrier than two single HD 5850, but the two HD 5870 pressed up 500 W the power usage of the system.