Are they some graphic card benchmark for linux environment ? From my windows experience, drivers are important, and often underestimate. My linux gaming experience is very bad, lots of my game are unstable, and others use a lot more resources than with windows. However, when I ask people, some of them have no issue at all, even with a similar environment (Debian + Steam). I may consider buy specific graphic card to stay on linux, but I couldn’t find any clue to know which one are more adapted.
Thx for your leads !
Which GPU do you have? Which drivers are you using? are you sure you’re using those drivers and they’re not just installed but unused? My first guess is that you have an Nvidia and are using open source drivers (nouveau).
Some performance difference is expected, after all most games are being run through a compatibility layer, and many others were ported as a second thought so they’re not optimized on the same level. Also note that lots of us don’t use Windows, so we’re not comparing experiences, if it runs at an acceptable frame rate with an acceptable graphics settings for what I would expect the GPU to be capable of, then I don’t bother benchmarking it.
Another consideration is whether they are plugged into the graphics card. Common performance “problems” arise when somebody tries to plug into the video-out on the motherboard, so they could be accidentally forcing the use of the iGPU, if present.
If using a somewhat modern distro, this isn’t an issue anymore (unless you run a really old OpenGL game).
I run my PC in this way with little to no performance degradation: monitors go to my motherboard (r5 2400g CPU with vega11 iGPU) and games use my RX 5700XT without having to do anything at all… Pretty smart handling tbh
That is, because supposedly that limitation still affects Windows. Do you use
supergfxctl?I’m running vanilla Fedora 40. Haven’t installed that, and just checked and it’s not even on fedora’s repos
I just remembered that that package isn’t compatible with Plasma 6 (yet), so maybe it got dropped from the official repos when they officially released 40.
True, but I don’t think it’s the case for OP since he reported less performance than on Windows, so I assume he meant on the same hardware.
I think that’s more of a proof that I shouldn’t answer these before I’ve had my caffeine. Good catch!