It seems we’re finally onto a heavy duty title, with Crysis Warhead. Crysis one, was one of my favourite games, however it had its flaws, such as memory leaks, more notably in the last level, and it also had some horrific performance problems. Thankfully, Crysis Warhead is a much better running game, thus why I’ve chosen to use it as the benchmark, as opposed the original. Another reason I’ve opted for Warhead as opposed the original, is that Warhead is a fair bit faster paced and pretty much throws you into the action.
In Crysis Warhead, I ran the game with everything set to enthusiast (Maximum), however I opted not to run any anti-aliasing, as simply, I don’t feel that Crysis at 1920x1080 benefits from it, and will only get a negative impact frame rate wise. Crysis Warhead is an absolutely stunning game to look at, and from the time it came out was a massive step forwards in term of graphics, with gameplay to match, with its open style gameplay in that you’re able to approach your objective from different locations.
For recording FPS with this game, as the other games, I ran a 10 minute sequence from when you take control of the character. In this sequence, it had a fair amount of explosions from vehicles, a nice FMV with lighting effects. Here are the results that I recorded;
Crysis Warhead - Results
We’ll start with the average FPS. I was actually quite surprised to see it, as what I consider quite high for Crysis Warhead, the game that would bring systems to their knees is now tamed by mid-range cards in Crossfire, while this will always become the case, it’s still a nice thing to see, as this is real gameplay, as opposed the fly over GPU benchmark that isn’t actually representative of gameplay. Looking at the minimums, they’re above 30 and ultimately I couldn’t see the game slowing down or stuttering in.
Dirt 3, the successor to the first DX11 title that was the spectacular looking Dirt 2, is the next game that I used, as a DX11 title I felt that it would be appropriate. I was originally going to choose Dirt 2, however I figured that I’d go with the successor, in the hope that they’ve implemented DX11 effects more noticeably, as opposed the subtle differences in the former. Alas, that isn’t the case, but it’s still a brilliant looking game and as good as Dirt 2 to play.
In Dirt 3 I was using 1920x1080 with the ultra preset and then set 8x MSAA, which looked absolutely brilliant and in my opinion is one of the better looking racing games that are on the market today. For recording FPS, I played through the first 10 minutes of gameplay starting at the first race, I felt that this was fair, as it allowed me to run through more than one different track to see how the different tracks would affect frame rates, here are the results;
Dirt 3 - Results
I was expecting the minimums to be slightly higher, but that could have been due to loading etc, not to fret, as with other games in this review, I didn’t notice any stuttering at all while playing, overall it was an extremely smooth experience, the averages are approaching 100, so again like other titles, could give headroom to increase resolution, overall the frame rate seemed very consistent throughout playing and the 6870 Crossfire proves it’s able to pump out the frames at maxed settings.