14 DAY TRIAL // Usually, video cards from AMD and NVIDIA covering the same price point are fairly similar in performance, but the case of the FirePro and Quadro professional boards is different for once.
The AMD FirePro V3900 has been pitted against an Nvidia Quadro 410, and the FirePro V4900 was compared to the Quadro K600 in SPECviewperf.
Normally, when we hear about benchmarks, we think about 3D Mark and other suites like it, and usually our expectations are met.
But this time around, it's not a comparison between consumer, gaming video cards that we're looking at.
Instead, the AMD FirePro series of professional graphics adapters has been pitted against the NVIDIA Quadro series.
Needless to say, a different benchmark was needed for this, and since SPECviewperf is endorsed by both AMD and NVIDIA, there aren't many alternatives that would be better.
Granted, these tests do seem to have been carried out by Advanced Micro Devices, but the specifications of the test benches seem to be identical, save for the video cards themselves, so we suppose they are as neutral as can be expected from a company that is always looking to one-up its rival.
Sapphire has a stake in this too, since it was its AMD FirePro cards that got tested. Here are the core components of the test computers: OS: 64-bit installation of Windows 7 SP1, Intel E5-1660 3.30GHz CPU, 16GB RAM, and these drives: AMD Catalyst 13.25.18.1 for the FirePro, Nvidia 331.82 for the Quadro.
SPECviewperf tests the real-world usage of cards, or emulates it anyway, and has viewsets for animation suites like CATIA, CREO, MAYA, SNX and Solidworks.
AMD's cards seem to totally trounce NVIDIA's similarly priced units, much to general shock. FirePro V3900 achieved 3.6 times better performance in the Maya test compared to an Nvidia Quadro 410, for example, and the FirePro V4900 beat the Quadro K600 in all eight tests. Check out the slides provided by VR-Zone for more.
Now we just have to wait and see how long it takes NVIDIA to provide some favorable test results of its own.
