Today we’re checking out Final Fantasy XV CPU performance using the new standalone benchmark released ahead of next month's PC launch. The reason we want to look at CPU performance first is because the game is extremely CPU intensive, far more so than we were expecting. Developer Square Enix recommends gamers pack at least a Core i7-3770 or FX-8350, that’s significantly more firepower from Intel, but we’ll look into that shortly.
The point here is that the game calls for an 8-thread CPU for the recommended specs, while on the GPU front, they believe you can get away with a GTX 760 or R9 280. Oddly in that second comparison AMD packs far more firepower, but then again this is an Nvidia sponsored title with GameWorks baked in.
According to Nvidia the game features integration of Nvidia Flow, Nvidia HairWorks, Hybrid Frustum Traced Shadows, Turf Effects, and Voxel Ambient Occlusion. As I understand it, these GameWorks features are only enabled with the High quality preset so although the CPU testing has been carried out using a GeForce GTX 1080 Ti, I’ve also tested with the Standard and Lite quality presets as well.
The standalone benchmark runs for over 5 minutes and covers several events, maps and characters used in the game. For our test though we are reporting frame rates across a 90 second piece of that benchmark and all results are based on an average of 3 runs. The test begins at the start of the car journey and ends shortly after everyone gets out of the car. Please note I found that frame times were often negatively impacted between the fade to and from transitions, so I avoided testing across those.
As usual, any auto overclocking functions in the BIOS were disabled and locked CPUs were paired with the appropriate memory, for example the Core i3-8100 was tested with DDR4-2400 memory. The unlocked processors used DDR4-3200 memory and this was also true for all the AMD Ryzen CPUs.
Starting things off we have the High quality results at 1080p. This is the maximum quality preset which has all the GameWorks features enabled. Here the GTX 1080 Ti created a system bottleneck with an average of 91 fps, so it’s quite shocking to see such a high-end GPU limited to less than 100 fps at this relatively low resolution.
For this graph let’s focus on the average frame rates before sorting the data by frame time results. Here we see that the Ryzen processors all perform very well in relation to Intel 8th gen CPUs, though Intel is being limited by the GTX 1080 Ti.
Even the much older Core i7-2600K and FX-8350 look quite good though the A12-9800 falls into a heap and isn’t able to deliver playable performance.
Sorting the data by the 1% low frame time results we see that the 8700K is 23% faster than the Ryzen 7 1800X and Ryzen 5 1600X. The 1800X is roughly on par with the old Sandy Bridge 2600K and quite shockingly the FX-8350. That said we see reasonable performance across the entire field with the exception of the A12-9800.
Then if we focus on the 0.1% data the older 2600K and FX-8350 processors fall away, as for the modern quad-cores such as the Ryzen 3 1300X and Core i3-8100, though the higher clocked 8350K still does quite well.
Since we were so heavily limited at 1080p with the GTX 1080 Ti, at least for the high-end 8th gen Core processors I decided to try 720p resolution. Here the 8700K is now 11% faster than the 1800X for the average frame rate, though the 1% low data was much the same despite pulling away by an 18% margin for the 0.1% low result.
There’s the high quality results and for the most part things looked pretty good. However because the results are based on an average of three runs they in a way smooth out some of the frame glitches we saw now and then.
Even with the Core i7-8700K and GTX 1080 Ti we still saw occasional lag spikes in the frame rate and this was really only a problem when using the high quality settings. If we disable GameWorks features using the standard quality preset performance smoothed out and we saw a few interesting things, let’s move on to take a look.
Here are the standard quality results at 1080p and right away we are seeing more consistent performance, with perhaps the exception of the older 2600K and FX-8350 processors.
However looking at the 8700K and 1800X we now see a much smaller difference in performance, the 8700K is now 12% faster when comparing the 0.1% results, whereas previously it was almost 30% faster. The variation in the average frame rate and 1% low result is also far less significant.
In fact, Ryzen looks very impressive using the standard quality settings as the Ryzen 5 1400 delivers much more consistent performance when compared to the Core i3-8100, for example.
If we sort the graph by the 1% low results we see just how close it is between the top 6 processors which includes the Ryzen 7 1800X along with the Ryzen 5 1600X and 1500X. Meanwhile, the Ryzen 3 1200 is comparable to the Core i3-8350K and Core i7-2600K which is a solid result. The FX-8350 also does surprisingly well getting amongst that bunch as well.
While I was at it, why not invest the time to check out the Lite quality preset and this lead to some very interesting findings...
Normally Intel’s higher clocked CPUs run away with low resolution, low quality testing as they can provide much higher maximum frame rates, and we do see that here when looking at the average frame rates.
The 8th Gen Core i5 and Core i7 processors were able to max out the GTX 1080 Ti at 171 fps making them at least 11% faster than the Ryzen 7 1800X. However you might have noticed something strange when looking at the frame time data, so let’s focus on that.
Here we find something unexpected. When sorting the data by the 1% low result the Ryzen CPUs come out on top, in fact even the Ryzen 3 1300X is able to beat the Core i7-8700K. That’s not the case for the 0.1% result, but here the Ryzen 5 and Ryzen 7 CPUs still beat the mighty Core i7-8700K. This is very unusual and we’ll have to dig a bit deeper to find out why we’re seeing this.
Waiting for the Actual Game
Those were some interesting results and should give you a good idea of the kind of processing power you’ll require to play Final Fantasy XV come March 6.
Normally we'd assume the benchmark tool is very accurate given the developer has released it ahead of time to allow gamers to prepare their systems. That said, evidence has surfaced that the tool might not be useful for measuring actual/expected GPU performance for the game. What's probably an unintentional bug in the benchmark, it renders objects it doesn't need (out of field/view, using GameWorks features) and is therefore overly taxing on the graphics card. So it might not be accurate for GPU testing, but we believe the CPU results should prove accurate.
Based on our testing, if you have a quad-core that’s clocked below 4 GHz you’re probably going to struggle with anything more than the Lite quality setting. The game is very CPU intensive. For example, the 12-threaded 8700K rarely dipped below 40% utilization and in our test spent most of its time over 50%, at times hitting as high as 80%. It was a similar story with the Ryzen 7 1800X though more time was spent hovering around the 30-40% mark. Still the lower-end Ryzen CPUs featuring 12 or less threads were very well utilized.
We should note the 1600X and 1800X were 5-10% faster with SMT disabled depending on the test, but I didn’t go to in depth there as we doubt many of you are going to disable SMT just to gain a bit of extra performance in Final Fantasy XV, but note you could if you wanted to.
The High quality settings which apparently enable all the GameWorks features were a bit concerning, stuttering was an issue here even with the most extreme hardware configurations. But it's possible this is due to the bug that renders out of view objects. On that note, AMD has said they won't release their optimized driver until next month when the game is out.
Obviously we're really keen to test the game once it’s released and hopefully we’ll have a great deal of settings that we can fine tune for better results. We'll evaluate doing a proper GPU benchmark test for Final Fantasy XV once we get closer to release and can confirm it will reflect the actual gameplay experience.
- Ad-free TechSpot experience while supporting our work
- NEW Get a free copy of Superhot: Mind Control Delete, courtesy of GOG
- Our promise: All reader contributions will go toward funding more content
- That means: More tech features, more benchmarks and analysis