S.T.A.L.K.E.R. 2: Heart of Chornobyl Update 2.0 Performance – Unreal Engine 5.5.4

Conclusion

There you go, Back To The Zone – Update 2.0 for S.T.A.L.K.E.R. 2: Heart of Chornobyl is out now. We thought it was quite exciting to see a big game engine update like this in a game, moving from Unreal Engine 5.1 to Unreal Engine 5.5. You don’t often see major game engine updates like this almost 2 years into the game’s release. Even more, GSC Game World went out of their way to improve the visuals along with this patch. There are real lighting, shadows, reflections, and foliage changes that you can notice straight away; it’s pretty impressive and moves this game closer to what they probably intended at launch. This is always how they wanted the game to be; it just needed time to get there.

In today’s S.T.A.L.K.E.R. 2: Heart of Chornobyl Patch 2.0 versus 1.7 Performance Review (Unreal Engine 5.1 vs Unreal Engine 5.5), we looked at some GPU-oriented gameplay to see what the performance differences are. We utilized native resolution and upscaling across 4K, 1440p, and 1080p. We also looked at some CPU utilization testing and VRAM testing, combined with two different types of run-throughs. This is but a taste; we originally wanted to cover all the GeForce RTX 50 series and Radeon RX 9000 series GPUs, but the task was not needed after we saw the pattern and understood what was happening. We did test a few more GPUs not shown here, and the pattern was exactly the same with them, so we can make some educated conclusions.

Performance

The first thing to grasp with the performance numbers is the fact that the visuals have changed between patches. Patch 2.0 improves the visuals, and it is noticeable. The foliage, for example, is much denser, more natural-looking, and certainly makes a greater demand on the performance. The lighting is also different, and noticeably looks better. These changes cannot happen for free, but we think the performance drain isn’t too severe for the increased visuals. In addition, it is really the CPU-oriented things, like towns/villages, that receive the most benefit in performance, but we are getting ahead of ourselves. We were using a Ryzen 7 9850X3D, which is going to give a better result; lower-end CPUs may scale even better with this update.

There was a clear pattern that faster GPUs like the GeForce RTX 5090 receive less of a hit to performance with Patch 2.0; in fact, they can receive a performance increase, at least on average FPS. But we found that as you descend in the GeForce RTX 50 series product stack, there was a consistent result of Patch 2.0 being slower than the previous patch. This got worse the farther down the stack you went. We also tested a GeForce RTX 5080, a GeForce RTX 5070 Ti, and a GeForce RTX 5060 Ti 16GB not shown in this review. However, each one was similar to the GeForce RTX 5070 results; Patch 2.0 was slower on average and in 1% Lows. The percentage differences varied, but it was common to see just under a 10% loss in performance across the board.

The pattern, though, was very different with the AMD Radeon RX 9000 series of GPUs. The Radeon RX 9070 XT performed exactly the same between both patch versions at 4K, except the 1% Lows were a bit lower. However, when we lowered the resolution to 1440p, the Radeon RX 9070 XT was actually faster with Patch 2.0! The Radeon RX 9060 XT 8GB was also very close in performance between the two patch versions. This shows that the Radeon RX 9000 series does not take as much of a hit to performance with Patch 2.0 as the GeForce RTX 50 series does. Consider that the visuals are improved, but the Radeon RX 9000 series absorbs this visual improvement, well, better.

When it came to evaluating the CPU-oriented performance, this was also interesting. The GeForce RTX 50 series had a larger advantage with Patch 2.0 in the CPU-oriented game areas. The Radeon RX 9000 series did receive a benefit with Patch 2.0 here as well, but to a lesser degree than the GeForce RTX 50 series. The GeForce RTX 50 series, basically, seems to be more sensitive to the CPU-oriented improvements. This is also backed up by the CPU utilization, which resulted in the CPU utilization being much higher on the GeForce RTX 5090 compared to the Radeon RX 9070 XT. In fact, with Patch 2.0, CPU utilization went up much higher on the GeForce RTX 5090 than it did on the Radeon RX 9070 XT.

Finally, we experienced a consistently higher VRAM dedicated usage with Patch 2.0. With the GeForce RTX 5090, it went up about 1GB higher, and with the Radeon RX 9070 XT, it was about 1.5GB higher. Interestingly, the Radeon RX 9070 XT consumed more VRAM and shot just over 12GB of usage. With longer playthroughs and more densely populated village areas, this could well exceed 12GB+. ASYNC Compute was also interesting; the GeForce RTX 5090 benefited way more with it, compared to the Radeon RX 9070 XT, at least at 4K.

It was interesting, though, that 1% Lows were consistently lower with Patch 2.0 across the board in our testing, hrm…

Final Points

Update 2.0 for S.T.A.L.K.E.R. 2: Heart of Chornobyl is a great and welcome update. We instantly noticed the visual improvements when we launched the game after patching. The foliage is great; it does demand higher performance, but it looks good. We were also impressed with the lighting. Previously, the lighting looked a little washed out or ‘grey’, but now the lighting looks better with contrast. We ran the game on Epic, but you can always turn game settings down to get better performance, of course, and the one I’d lower first if you need better performance for GPU-oriented areas would be the foliage.

This is also a game that will benefit from using upscaling, and give you much more playable performance. This game now natively supports FSR 4, and you can also enable DLSS 4.5 and different models from the NVIDIA app. The game also natively supports AMD’s and NVIDIA’s best latency technologies, which helps. The one thing this patch did not add that would have been interesting would have been some Ray Tracing features. Oh, and that darned “Shading Compilation” loading screen. That has been the one thorn in the side with this game; it takes a long time, and it happens at every launch. That did not get fixed or changed with this update, and I was really hoping it would.

I would definitely say the visual upgrades warrant the performance results we experienced; it is acceptable. There are real benefits in the areas that were tougher to run, like towns/villages, and it will probably scale better on lower-end CPUs. Otherwise, this new patch now gives me reason to finally finish the game. I like enjoying the game as the game developer intended it to be, and I think this brings it closer to their vision.

Join the discussion in The FPS Review Forums...