I'm playing ARK: Survival Ascended in a huge base with tens of thousands of structures loaded at once. I usually get around 20–30 FPS depending on my location, and performance can drop below 10 FPS, even though my RTX 5080 stays at only 20–25% utilization. My system has a Ryzen 9 7900X, RTX 5080, MSI PRO B650-VC motherboard, 32GB of DDR5-6000, and a WD SN8100 SSD. All of my graphics settings are already very low, with an aggressive configuration that removes or reduces effects such as fog, clouds, grass, weather, water, and textures. Some players claim to get roughly twice my FPS with similar or weaker systems, but with CPUs such as the 7600X3D. Would upgrading from the 7900X to something like a 9800X3D make a major difference in a situation this heavily limited by the CPU, or is there likely another problem with my system or game configuration? EXPO is enabled at DDR5-6000, but I haven't successfully overclocked or tuned the CPU yet.
4 Answers
The 7900X is not a bad gaming CPU, but the X3D models can be substantially faster in CPU-bound games. The likely benefit is better frame rates and especially better 1% lows, not automatically turning 10 FPS into 100 FPS. If your current processor is healthy and the game is genuinely limited by its main thread, a 9800X3D is the sensible upgrade to consider, but try to reproduce the other players’ settings and test in the same part of the base first so you know the upgrade is addressing the real bottleneck.
Before buying a CPU, verify that the 7900X is boosting normally and that nothing is throttling it. Check temperatures, effective clock speeds, BIOS settings, chipset drivers, background processes, and memory stability. DDR5-6000 with EXPO is a good starting point, but make sure the system is actually running dual-channel and that the game is not using a problematic launch option or configuration tweak. If reducing settings has almost no effect and one game thread is maxed out, the engine is probably the limiting factor.
X3D chips can help a lot in games that constantly revisit large amounts of world or simulation data, especially when the CPU is limiting the GPU. However, the improvement is very game- and scenario-dependent. Average benchmark results often understate the benefit in an unusually large late-game base, but a consistent two- or three-times increase from changing only the CPU would still be unusual. A 9800X3D would be the most straightforward upgrade for gaming on this platform; the 7900X3D and 7950X3D are less attractive unless you specifically need their extra cores.
That’s what I’m trying to determine. Most reviews test normal saves, not a base with tens of thousands of structures, so I’m wondering whether the extra cache behaves differently once the game becomes this CPU-limited.
The low GPU usage confirms that the graphics card is waiting on something, but it doesn’t prove that average CPU performance is the only issue. Check per-thread CPU usage, frametimes, and whether one or two cores are saturated; total CPU usage can look moderate while the game thread is completely maxed out. Also compare the exact same location, camera direction, save, resolution, and configuration when testing other systems. ARK’s late-game simulation and rendering workload can vary enormously between bases.
The other players may also have different base layouts, creature counts, server settings, BIOS versions, or configuration files. A direct side-by-side test in the same area would be much more useful than comparing reported FPS.
Everything is already set to very low, including a heavily reduced configuration file, and EXPO is running at 6000. I haven’t managed to overclock the 7900X because the guides I found use different BIOS menus, but I’ll check boost clocks and temperatures before replacing anything.