
Linux vs Windows Benchmark MIDDLE EARTH SHADOW OF WAR
MIDDLE EARTH SHADOW OF WAR has been tested in Linux (Pop!_OS 24.04, COSMIC/Wayland) and Windows on my dual-boot machine:
RTX 5070 Ti
Ryzen 9 5900X,
RAM 32 GB
Each operating system has its own identical 1TB SSD drive.
The game was run at 1080p using the Ultra settings, DLSS OFF, VSYNC OFF, FRAME LIMIT OFF.
The battle was extremely close, according to the results. Linux produces stronger 1% lows by about 10 FPS, resulting in slightly superior stability during heavy sequences, although Windows leads by about 5 FPS on average. Both platforms work quite well, often reaching 200+ FPS, which results in incredibly responsive and fluid gameplay.
Neither system had any stutters, stability issues, or visual discrepancies. This benchmark, which shows sophisticated optimization where Linux and Windows now compete nearly head-to-head, is a great illustration of how well DX11 games perform with Proton.
I'm still looking forward to the next NVIDIA driver 595 update for POP_OS, which should boost Linux DX12 game performance.
***********************************************
Disclaimer: Why I Test with Pop!_OS + NVIDIA
***********************************************
- Windows gamers
The whole point of these benchmarks is to show that Linux gaming exists, works well, and isn’t nearly as complicated as many Windows users think. I’m basically trying to show a realistic migration path from Windows to Linux, not build a perfect Linux-only lab.
- NVIDIA dominates the gaming GPU market.
According to the Steam Hardware Survey, NVIDIA usually sits around ~75–80% of GPUs in gaming PCs. If I test on NVIDIA, I’m covering what most gamers actually use.
- Pop!_OS is one of the easiest distros for NVIDIA users.
It ships with dedicated NVIDIA ISOs, drivers are integrated, and updates are straightforward. I run tests on official Pop!_OS drivers, so the setup reflects something an average user could realistically install.
- If Linux gaming works on NVIDIA, it works for most gamers.
Yes, AMD often performs better on Linux. I’m aware of that. But testing only on AMD would shrink the scope from ~80% of the market to a much smaller slice. My goal is broader relevance, not best-case scenarios.