For years people debate which video cards are better. If apparently they are very close at each generation, practically the Radeon architecture is many times faster and better.
Why they appear close ?
Well, the main reason for that is software developer’s lazy mode.
All games are being launched fast and simple instead of actually being optimized for advanced architectures.
A secondary reason would be that nVidia bribed (oh…sponsored) way too many of the game companies to optimize their games to favor nV hardware. That being done, yes, the benchmark results get pretty close.
But what you should know is the real processing power of a card. That’s measured now-days in Terra floating point operations per second.
Just looking at the official specs, you will see that Radeons have significantly more TFlops than GeForce cards, therefore Radeons significantly more processing power.
An already old Radeon 5870 has 2.72 TFlops compared to the supposedly superior GTX 295 card that has only close to 1.79 TFlops.
Comparing any 2 same-generation cards will prove beyond any doubt that Radeons are by far better when it comes to processing power.
The current top cards, Radeon 7990 and GTX 690 have still significant TFlops difference. While the Radeon has 8.2 TFlops, the GTX has only 5.6. They (nVidia) don’t even have the decency to list it on their site.
That’s being said from the tech perspective, the values differ even more in practice.
But to use the complete capabilities of those cards, you actually need software that can do that.
Such software are the ones used for mining Energy Credits like bitcoin. They process hashes.
On average, the same generation Radeons are 4 times faster then the GeForce cards, getting even close to 8 times faster in certain models.
The best desktop Radeon exceeds 800, while the best desktop GeForce barely hits 160. M Hashes/s.
A very large table of hardware can be checked here:
So there you have it, once and for all, Radeons win hands down, for any generation.