For our gaming tests, we’ll be testing three games: Tomb Raider (2013), Crysis 3 and Middle Earth: Shadow of Mordor. We’ll be testing all of them at 1080p, as that is the most common resolution for PC gaming. A lot of sites will test using resolutions like 800×600 or 1024×768, but as this does not represent the vast majority of user’s configurations we feel that it provides synthetic results, and we prefer to keep our gaming benchmarks to real-world situations. We’ll also be testing gaming while streaming, which we feel is very important, especially with a platform like X99.
FRAME TIME VARIANCE
For measuring gaming performance we’ll be employing both standard FPS tests as well as frame time variance, which allows us to further analyze the ‘smoothness’ of gameplay. As a benchmarking tool, frame time is as important if not more so than frames per second as it allows to us to see more than FPS can provide alone, like micro-stutter.
So, what exactly is “frame time”? To put it simply, it is the measurement of the amount of time it takes for a frame to be drawn, the lower the amount the smoother your experience will be. What the limit is before you’ll notice there’s an issue is going to vary from person to person, but generally anything below 30ms (milliseconds) should be acceptable. We’ll be measuring both the average frame time and the 99th percentile, which is essentially the worst results you can expect 99% of the time.
We tested on “High” settings at 1080p with FXAA.
As we can see above, with standard FPS testing nearly all the configurations performed about the same, with the exception of the average FPS being slightly higher on our 4.5GHz results, however.. When we look at frame time, we can see that while all three configurations performed quite well, our 4.5GHz overclocked does a bit better.
SHADOW OF MORDOR
We tested Shadow of Mordor on “Very High” at 1080p.
Not only do all three configurations play this game at nearly the same exact FPS, but the frame times are identical. It’s quite astounding, really.
TOMB RAIDER 2013
For Tomb Raider, we tested the game on “Ultimate” settings with TressFX and FXAA turned on at 1080p.
Here we can see that once again our results are very similar across all three configurations, with our stock CPU getting a slightly higher average FPS, but at 1.25 frames, it is within the margin of error. Also, oddly enough our stock clocked chip seems to have fared slightly better in frame time as well.
We tested streaming using OBS (Open Broadcaster Software) with the recommended settings for Twitch streaming, 1080p (downscaled to 720p) 60FPS. Our in-game settings remained the same at “Ultimate” with TressFX and FXAA at 1080p.
We can see that compared to our non-streaming test, we seem to lose an average of ~4 FPS while streaming. This is fairly good considering how taxing encoding video on the fly can be.