We would like to thank the following companies for supplying and supporting us with our test system hardware and equipment: Intel, ASUS and Corsair.

As far as the testbed goes there are no surprises as usual. We've got our typical X79 setup that we've been using since the launch of Sandy Bridge-E. As for the video cards we'll be using today we've of course got the new GTX 670 at its reference clock speeds which we will discuss in just a minute.

We've also got the older GTX 580 and the original GTX 680. Because that was on a slightly older driver we've also included the GTX 680 4GB from Palit which will give us a better idea of stock performance with the most recent driver as we know the extra memory doesn't make too much of a difference at most resolutions.

Along with those cards we've got the HD 7950 and HD 7970 which is of course the main competition to the new GTX 600 series cards along with the older dual GPU HD 6990 from AMD to round out the selection today.

Before we get into the testing side of things, though, we want to just take a quick look at the specifications of the video card here today. Looking below we can see there's 1344 Unified Shaders present, this is slightly down on the 1536 offered by the GTX 680. We can see 32 ROPs are present like the GTX 680 and the memory bus also lines up coming in at 256-bit.

As for clocks the core comes in slightly lower at 915MHz while the 2GB of GDDR5 actually carries the same 6008MHz QDR clock as its big brother. For the most part there's not a whole lot of difference between the two cards. What will separate them is the Shaders ultimately. As you can see in the GPU-Z screenshot above, we blocked out some information to protect the company who sent us the video card.

Let's get started!

The FPS Numbers Explained

When we benchmark our video cards and look at the graphs, we aim to get to a certain level of FPS which we consider playable. While many may argue that the human eye can't see over 24 FPS or 30 FPS, any true gamer will tell you that as we climb higher in Frames Per Seconds (FPS), the overall gameplay feels smoother. There are three numbers we're looking out for when it comes to our benchmarks.

30 FPS - It's the minimum number we aim for when it comes to games. If you're not dropping below 30 FPS during games, you're going to have a nice and smooth gaming experience. The ideal situation is that even in a heavy fire fight, the minimum stays above 30 FPS making sure that you can continue to aim easily or turn the corner with no dramas.

60 FPS - It's the average we look for when we don't have a minimum coming at us. If we're getting an average of 60 FPS, we should have a minimum of 30 FPS or better and as mentioned above, it means we've got some smooth game play happening.

120 FPS - The new number that we've been hunting down over recent months. If you're the owner of a 120 Hz monitor, to get the most out of it you want to get around the 120 FPS mark. Moving from 60 FPS / 60 Hz to 120 FPS / 120 Hz brings with it a certain fluidity that can't really be explained, but instead has to be experienced. Of course, if you're buying a 120 Hz monitor to take advantage of 3D, an average of 120 FPS in our benchmark means that in 3D you will have an average of 60 FPS, which again means you should expect some smooth gameplay.