Passmark graphics benchmarks: accurate?

alpha_c

Distinguished
Jul 10, 2010
29
0
18,540
http://www.videocardbenchmark.net/video_lookup.php?cpu=FirePro+V4800

The Firepro V4800 ranks 37, the Quadro FX 1800 ranks 94 and is about 3x the cost...
The Firepro is supposed to be a Radeon HD 5670 core. The HD 5670 ranks 34, the Firepro V4800 ranks 37. This suggests that drivers have much impact on these benchmarks (dropping it from 1165 to 1134).

This rank is also higher than the Quadro FX 5600 and Quadro FX 3800 (both of which are multiples of the Firepro's cost).

Similarly, the Firepro V5700 ranks higher than the Quadro FX 1800. It is based on the Radeon HD 3800 series and it takes a hit from rank 51 (HD 3850) or 77 (HD 3870) to 85. The difference in the Passmark scores is from the HD 3850's score of 810 to the V5700's score of 768.

Interestingly enough, the Firepro V3800 based on the Radeon HD 5570 takes a huge hit. The Radeon HD 5570 is rank 84 while the Firepro V3800 is 164... (from 744 to 490)

Are these benchmarks accurate at all, given their reliance on drivers? (There is no motherboard, power supply, or processor "control" so to speak)
 

ubronan

Distinguished
May 13, 2009
42
5
18,545
I dont like this benchmark program, there are many better onces out there.
What i miss is some real references to similar systems which i could not find.

3dmark ofcourse leading with their products like pcmark, if its just about graphics furmark is a good way to see how it performs with its benchmark.

Its also a super toy for finding the best oc on you videocards, it surely know how to heat up your vc a lot xD