Hey there. Over the past two days I ran multiple gaming benchmarks on the "blazingly fast" CachyOS and openSUSE's immutable flagship distribution Aeon.

The initial reason as of why I did this was because I read over the past months quite a few articles about CachyOS and how it delivers x86_v3 or even x86_v4 optimized packages. Being Arch Linux based and therefore rolling.
While nobody seemed to acknowledged that openSUSE Tumbleweed and especially openSUSE Aeon do the same thing on eligible hardware.
Therefore I did me own benchmarks to see if there are any noticeable performance differences between these two distributions.

The results are quiet interesting. tl;dr: No there are no note worthy performance differences in general. But let's dive deeper into the topic.

Hardware:
CPU: AMD Ryzen 7800X3D
GPU: nVidia GeForce RTX 3080
RAM: 16GB DDR-5 4800
Display: 2560x1440
Board: Gigabyte B650 GAMING X AX V2
SSDs: Samsung 860 EV 1TB + WesternDigital Blue 1TB (WDC WDS100T1B0AWDS100T1B0A)

The Steam Library was on the above mentioned SSDs. Both are ONE BtrFS file system. Which means files are evenly distributed across both devices. Which in theory increases loading performance and also limit the stress put on each device individually as they split the work.
Downside: If one device get broken all files are probably lost. But that what we have backups for. Well in case of the Steam Library I do not backup that one apparently.
Anyway, both OSes used the exact same hardware and file system to run the games from.

Also Shader Pre-Caching was disabled

Software:
openSUSE Aeon:
- Kernel 6.9.1-1-default
- Driver: 555.42.02

Note: For openSUSE Aeon I ran all benchmark from flatpak Steam and from the RPM steam via a Tumbleweed distrobox container. Tumbleweed is what Aeon is based on btw. Or in simple words Aeon is immutable Gnome Tumbleweed.

CachyOS:
- Kernel 6.9.1-4-cachyos
- Driver: 555.42.02

Games:

- Ashes of the Singularity: Escalation (GPU and CPU benchmark)
- Borderlands 3
- The Calisto Protocol
- Horizon: Zero Dawn
- Returnal
- Shadow of the Tomb Raider

Benchmarks:

(Right-click open image in new tab to enlarge)

Ashes of the Singularity: Escalation:
!Ashes of the Singularity: Escalation - CPU Benchmark Results
!Ashes of the Singularity: Escalation - GPU Benchmark Results

Borderlands 3:
!Borderlands 3 - Benchmark Results

The Calisto Protocol:
!The Calisto Protocol - Benchmark Results

Horizon: Zero Dawn:
!Horizon: Zero Dawn - Benchmark Results

Returnal:
!Returnal - Benchmark Results

Shadow of the Tomb Raider:
!Shadow of the Tomb Raider - Benchmark Results

Summary:

!Summary

Conclusion:
After all these tests it is safe to say that in general there seems to be no difference if someone uses CachyOS or openSUSE Aeon / Tumbleweed performance wise. At least for gaming.

Both deliver similar performance even thought Aeon scored the best (and worst in case of flatpak) in average. But we talk about a 2 - 4 fps difference here which in my opinion is neglectable.

Either both are making similar performance improvements, as both CachyOS and Tumbleweed/ Aeon ship at least v3 optimized packages or gaming is the wrong use-case to see any real performance benefits with x86 optimized packages.
After all both do run Proton and DXVK / VKD3D compiled and shipped by Valve. Except for Shadow of the Tomb Raider which was the only native Linux game I tested. But also it was compiled by Feral Interactive and just happened to be executed on the above systems.

The more interesting thing about these benchmarks is, that using flatpak does also not impact the performance in a notable way. Unlike anticipated by several ppl among the Linux Gaming community (see reddit ).
Only for Horizon: Zero Dawn it was unable to get above 86fps no matter how often I ran the benchmark. Which was a performance penalty of aprox. 20% here. I did not yet figured out as of why it happened for this particular game. Especially since it made no difference in any other game at all.

Again we talk about a difference of max 4 frames per second.
Looking in the performance differences as calculated in the summary looks more drastic than it is. I mean if you get an avg. of 100fps on system A and 104 fps on system B it is a 4% difference. Even though it is not a real difference at all.

Anyhow. Any thoughts on this one? Maybe I am drawing the wrong conclusions? Or my Benchmarking might be flawed?

Video:
https://youtu.be/AB4xhojV1sw

Last edited by Vortex_Acherontic on 29 May 2024 at 3:14 pm UTC