I want to improve gaming performance on Ubuntu without blindly copying a huge list of tweaks. What is worth trying, and how should I compare it with an Arch-based system?
How can I tune Ubuntu for gaming, and is Arch actually faster?
Answers
Start with a baseline before changing anything. Pick a few games you actually play, use the same resolution and graphics settings, and repeat the same test scene several times. Record average frame rate and, if your tool reports it, frame times or low frame rates as well.
Then change one thing at a time and repeat the tests. A result from one machine or one game doesn't establish that a tweak will help yours. For an Ubuntu versus Arch comparison, keep the hardware, game settings, driver versions, and test conditions as close as you can. Otherwise you may be measuring a different driver or game update rather than the distribution.
Reply to this answer
Don't paste a full boot parameter line just because someone calls it a gaming setup. Options such as noht, amd_pstate and the AMD GPU feature mask affect different hardware or behavior, and some may not apply to your system at all.
The GPU feature mask in that example is intended to expose AMD GPU overclocking features, not to speed up every game by itself. Check what an option does, whether your kernel and hardware support it, and how to undo it before testing. Keep a note of your original settings so you can revert a change that causes trouble.
Reply to this answer
Filesystem tweaks need context too. The example uses Btrfs mount options for compression, write timing, and access timestamps. Those aren't a drop-in recipe for an ext4 installation, and changing write timing can have consequences if the system loses power.
Same goes for the memory and writeback settings in the sysctl list. They are choices for a particular setup, not proven universal gaming defaults. I would leave them alone unless there's a specific problem to investigate and a way to measure the effect.
Reply to this answer
Ubuntu is a reasonable place to start. Try the games first, measure any problems, and only then decide whether a different setup is worth testing. A long optimization guide can be useful for ideas, but its title or distro label isn't evidence that every setting will help.
Reply to this answer
Add your answer
Answers are reviewed before publication.
raidTauntAfterBrunch:
That makes sense. I was mostly wondering whether switching distributions alone would be a big upgrade.
rogueSkillTreeJanitor:
The supplied comparison suggests Ubuntu can be competitive after tuning, but it doesn't establish a general performance gap or guarantee the same results on other hardware.