Zombie Army 4: Dead War benchmarks, settings, and performance analysis

Rebellion is back with another zombie horde shooter, Zombie Army 4: Dead War—a decent game that doesn't really shake things up in the genre. Much like Left 4 Dead, Strange Brigade, World War Z, and the previous Zombie Army installments, players are tasked with facing down the zombie apocalypse, though in the case of Zombie Army it's Hitler behind the undead, and you're encouraged to snipe off their heads with bonus multipliers—complete with Sniper Elite's zoom camera showing the path of destruction your bullet takes. But what sort of PC hardware does it take to run Zombie Army 4, and how do the settings affect performance and visuals?

Zombie Army 4 uses the same engine as Rebellion's Sniper Elite 4, which immediately feels a bit odd as Strange Brigade already did that in 2018, but you can never have too many zombies to kill. We've been using Strange Brigade as a benchmark for GPUs for the past year or so, and we're very familiar with how it runs. Zombie Army 4 is mostly similar—for example, it's nowhere near as demanding as Metro Exodus, Battlefield 5, or Call of Duty: Modern Warfare (2019). That's not a bad thing, as it's more accessible to lesser PCs, but it does run a bit slower than Strange Brigade—and perhaps looks better as well, though we're comparing zombified apples and oranges. 

Much like past games from Rebellion, Zombie Army 4 ticks most of the important boxes. Uncapped framerates, lots of resolution options including ultrawide and multi-monitor—with correct aspect ratio support—and plenty of graphics settings to tinker with. About the only thing really missing is mod support, which is sadly common with most major games these days. Manual FOV adjustments are also missing, though the game does properly extend the FOV based on your resolution. Zombie Army 4 comes with four presets, the choice of two APIs—Vulkan or DirectX 12—and 12 primary settings you can adjust, in addition to things like resolution, vsync, render scale, and brightness.

 Zombie Army 4 settings and performance

For our analysis of the various settings and how much they affect performance, we've started with the ultra preset, including enabling AMD's FidelityFX, and then tested performance with each setting turned down to minimum or off to measure the change in performance. We've included screenshots to help illustrate the difference in visual fidelity.

DirectX 12 or Vulkan: In limited testing, the choice of which API to use isn't always clear. We checked several GPUs, including an older GTX 980 card, and Nvidia GPUs seem to perform a bit better with Vulkan compared to DirectX 12. AMD GPUs on the other hand may favor DirectX 12 slightly. The difference is 3 percent or less, and things could change with further updates. You can easily check performance on your own PC with both APIs, and choose whichever works better, though long-term we expect Vulkan may win out.

Graphics Detail: The one-stop shop to adjust all the advanced graphics settings covered below, the can be set to low, medium, high, or ultra. High looks nearly the same as ultra and runs 12 percent faster on Nvidia and 20 percent faster on AMD. Medium starts to introduce more noticeable differences in image quality, but performance improves by nearly 30 percent compared to ultra. Finally, using the low preset drops everything to minimum quality and boosts fps by about 60 percent. The above gallery shows the four different presets.

Texture Detail: Lower resolution textures use less VRAM at the cost of fine details. Most GPUs with 4GB or more VRAM should be able to run ultra quality textures, but if you're looking for every bit of performance, dropping textures to low only improved performance by 2-3 percent.

Image 1 of 2

Shadow Detail: Shadow maps require VRAM and additional calculations by the GPU to generate correct shadows, and end up being the single most demanding setting. Turning shadow detail to low can boost performance by 15 percent, at the cost of the world looking less moody. We'd suggest going for a setting of medium or high if your hardware can manage that.

Reflection Detail: Reflection calculations can be quite heavy, though there aren't a ton of reflective surfaces in Zombie Army 4. Here we noted a much larger difference between the AMD and Nvidia GPU we checked. The 2060 only gained 6 percent with reflections at low, while the 5700 performance improved by 13 percent. Regardless, you still get some reflections, and you probably won't even notice the slightly nicer looking puddles.

Anti-Aliasing: Zombie Army 4 appears to use temporal anti-aliasing, which can introduce some fuzziness, though this is mostly counteracted by FidelityFX (see below). Disabling AA completely will result in some visible jaggies, and only improves performance by 3 percent.

Draw Distance: At least in the benchmark sequence, it's difficult to spot any major differences between low and ultra draw distance. That isn't true of all areas in the game, but draw distance is the second most demanding setting, and dropping to low can boost framerates by 11 percent. We also noted minimum framerate improved by 20 percent.

Image 1 of 4

Ambient Occlusion: This helps approximate shadows you see in corners, and helps make the game world look less 'flat.' On the GPUs we tested, turning AO off only improved performance by 3 percent on Nvidia, and 6 percent on AMD.

Motion Blur: Some people think motion blur looks great, others dislike the effect. Either way, this has a negligible impact on performance.

Obscurance Fields: The game says this "will render ray-traced soft shadows around the player and other characters," though it's important to note this isn't the same sort of ray tracing that Nvidia talks about with it's RTX graphics cards. Instead, this is an approximation that might better be called screen space ray tracing (SSRT), and it's very lightweight. Turning this off improves performance by only 2-3 percent.

Tessellation: Tessellation normally adds additional polygons (geometry) to flat objects, especially as you get closer to the surfaces of the objects. However, looking at screenshots, we couldn't spot any noticeable difference between having this on or off. Performance was basically unchanged on Nvidia GPUs, and just 2 percent faster on AMD GPUs.

Image 1 of 20

Use Async Compute: This feature can help optimize the utilization of GPU resources, particularly on AMD hardware. We didn't check older Nvidia GPUs, but at least with the RTX 2060 it's also best to leave this enabled. Turning this off drops performance by 3-5 percent on current generation GPUs, though older Nvidia hardware may gain a few fps with this disabled.

FidelityFX Sharpening: The antidote to temporal anti-aliasing, AMD worked with the creator of FXAA (fast approximate anti-aliasing) to create a contrast aware sharpening filter. We like the added sharpness it adds, and anti-aliased lines still look correct. Best of all, this is extremely low impact—turning FidelityFX off did not improve framerates.

Reduce Mouse Lag: Measuring whether this really reduces mouse lag is more difficult, but we can say for certain that this doesn't help with framerates. But it also doesn't hurt, so you can leave it enabled.

 Zombie Army 4 benchmarks 

GPU benchmarks

Your graphics card will be the biggest factor in determining how well Zombie Army 4 runs. Budget cards like the GTX 1650 are able to handle 1080p and medium settings at 60 fps, but you'll need something like a GTX 1660 or RX 590 to do 1080p ultra at 60 fps. Overall, however, this isn't a super demanding game. The RTX 2080 Ti breaks 144 fps at 1440p ultra, and comes close to 90 fps at 4K ultra. Using a blend of high and medium settings should allow plenty of other GPUs to hit 60 fps at 4K as well, if that's what you're after, and 1440p even at maximum quality runs well on mid-range GPUs like the RX 5700 and RTX 2060. 

Image 1 of 6

CPU benchmarks

CPU demands are far lower than GPU, and the difference between platforms and motherboards can have an effect on performance. We tested the Ryzen 3950X using MSI's X570 Godlike motherboard, and used the MSI X470 Gaming M7 AC for the Ryzen 5 2600. The previous gen Ryzen CPUs are normally quite a bit slower than third gen Ryzen, but the 2600 basically ties the 3950X. Obviously, core and thread counts aren't a major factor here.

The same holds for Intel CPUs. Sure, the i9-9900KS (or an overclocked 9900K) takes top honors when paired with an RTX 2080 Ti, but it's only 10 percent faster than the i3-8100 at 1080p low. The gap shrinks to 5 percent between the fastest and slowest CPU as graphics quality and resolution increase. That's not including the 3950X DX12 results, which are merely a point of reference. As noted earlier, Vulkan performs better with Nvidia GPUs by a few percent.

Image 1 of 6

Zombie Army 4 is our first settings and performance article for 2020, and in a nice change of pace, it's not quite as demanding as many of the games we tested last year. That respite isn't likely to last, however, with ray tracing games like Cyberpunk 2077, Vampire the Masquerade: Bloodlines 2, Watchdogs Legions, and more slated to release this year.