AMD Radeon Linux Gaming Performance At Parity Between KDE Plasma 6.0 X11 vs. Wayland

[qarium]

To be honest at this point I at least expected to have slighty better performance in wayland. Not by a lot, like 1-2%, instead it is slighty worse. Hopefully up-to-date gnome doesn't have performance regression in Wayland like few examples shown.
What I really like is KDE expierience is very consistent. There is no ups or downs between x11 and wayland. It gives some confidence that KDE6 Wayland is matured and you shouldn't expect something not working or working worse due to Wayland.
I think testing such thing makes more sense in opensuse tumbleweed or something like Arch. You get up to date both KDE and Gnome expierience.

what do you expect of Xwayland ? it is a legacy codepath... wait until Wine/Proton has wayland support then you see all the windows games in wayland mode and only then you see faster performance also keep in mind that the change from implicit sync to explicit sync also ingrease the performance a little.

[t1r0nama]

Don't care, tiling WMs are the spawn of satan anyway. They don't deserve to exist.

WTH? They are the most efficient way to use displays. In fact they are so efficient that even mac os and windows try to have same features as WMs on linux.

There is just small learning curve about 1-2 weeks and you'll get used to it and it will be harder for you to revert to DEs then it was switching to WMs.

[mSparks]

no serious gnome user is on this old gnome version... i am on fedora 40 and gnome 46 and i can not experience any of these slowdowns.

Its probably the last one that worked.

this test was bias from start to end. also why do you think that benchmark x11 vs Xwayland makes any sense ?

The posit was that X11 is slow, badly written, and to hard to develop for, and that wayland/xwayland would therefore far outperform it by stripping out all the "legacy" stuff and writing all the code with an emphasis on performance and "modern" C++.

just wait until wine/proton has native wayland support then all the windows games run native wayland instead of Xwayland.

I thought it already did, although Ive not used any windows stuff since windows 7 - do all my "gaming" on a PS4, save CS2, planning on getting a PS5/PSVR2 once Aces of Thunder drops if it is as good as it looks (along with granturismo of course)

My PSVR headset has been plugged into the linux workstation for most of its life now, almost entirely for flight training, I would already have a PSVR2, but with no PC/Linux support I've not bothered so far.

[oiaohm]

Seriously, Nvidia wouldn't be passing OpenGL validation tests (especially modern ones) if they failed such basic things. And yes people other than nvidia themselves were running them and they do pass fully 4.6

https://forums.developer.nvidia.com/t/piglit-opengl-conformance-test-seems-to-really-give-nvidia-drivers-grief/44967/2

Turns out when you go and double check Nvidia claimed opengl support things go south quickly. Lot of ways Nvidia like Volkswagen before emissions scandal. They use work around to make themselves past and when you test without the work around they are not up to standard.

[smitty3268]

you get the point that it was a 2 year old gnome version ? i run fedora 40 with gnome46
this means its pretty sure that no real gnome user would experience this ....

it is not anecdotal story its just the fact that a real gnome user just use fedora 39 or fedora 40... ubuntu is just slow here to update gnome.

At the very least, Michael posted benchmarks of Gnome 44 and the GravityMark vulkan benchmark there looked fine. He didn't show the CS2 test, unfortunately, though CSGO and Left 4 Dead 2 looked fine.

It did have a severe regression in DiRT Rally 2.0. Hopefully that's already been fixed. I don't think he's posted any Gnome 46 benchmarks yet.

[qarium]

Wayland offers no such functionality, non wayland devs have put some effort to hack it in, and they all do it differently, to various degrees of success.
X11 offers this functionality explicitly by means of Linux uid r/w/x permissions to the underlying bytestreams.
Well, there is at least one important difference, about 30 years of development time, resolving security issues/requirements and bug fixing.

time is not a relevant metric in the open-source software world we use a lot of unix standards much older...

of course time in the closed source world is much more relevant and companies like microsoft and apple have great pressure to bring up new stuff in short period of time just to make sure they can sell new stuff.

who cares if x11 is 30 years old and wayland 15 years old ? i can say for sure i do no longer use x11 at all.

[qarium]

That was absolutely the plan,
A plan that assumed they would stop taking new X11 contracts in 418 days from today.
A plan that assumed widespread wayland 3PD adoption well before now due to its far superior performance and being much easier to develop for and support
A plan that assumed only RHEL contributes to Linux desktop
A plan that assumed making wayland default would make people use it even though it didn't work
I think they are probably going to need a new plan.

this managment dreamland of plans and strategic future planing well my opinion is it will never work. its all wishful thinking.
and outside of the high-managment filter-bubble honestly no one believe in it. i did read some IT-Managment books and its full of this bullshit.

for a manager all is like central planing economics like UDSSR there is litterally no difference between high-managment and socialism/communism.

socialism/communism is the illusion that you can manage a society by central planing ..

[qarium]

Its probably the last one that worked.
The posit was that X11 is slow, badly written, and to hard to develop for, and that wayland/xwayland would therefore far outperform it by stripping out all the "legacy" stuff and writing all the code with an emphasis on performance and "modern" C++.
I thought it already did, although Ive not used any windows stuff since windows 7 - do all my "gaming" on a PS4, save CS2, planning on getting a PS5/PSVR2 once Aces of Thunder drops if it is as good as it looks (along with granturismo of course)
My PSVR headset has been plugged into the linux workstation for most of its life now, almost entirely for flight training, I would already have a PSVR2, but with no PC/Linux support I've not bothered so far.

no absolutly no one ever expected Xwayland to outperform X11... the only expectation in performance was native-wayland should outperform x11.

Xwayland is legacy only code. no sane person should ever be interested in legacy-only code

"and to hard to develop for"

no one ever expected a easy-to-develop-for solution because they did know that as soon as they want to increase security it will automatically become hard-to-develop-for.

any person who thinks you get security for free and this "easy" is an insane person.,

[qarium]

At the very least, Michael posted benchmarks of Gnome 44 and the GravityMark vulkan benchmark there looked fine. He didn't show the CS2 test, unfortunately, though CSGO and Left 4 Dead 2 looked fine.
It did have a severe regression in DiRT Rally 2.0. Hopefully that's already been fixed. I don't think he's posted any Gnome 46 benchmarks yet.

i can say just the switch from gnome45 to gnome46 and it feels much faster/snappier... of course only feeling no benchmark numbers.

[mSparks]

time is not a relevant metric in the open-source software world we use a lot of unix standards much older...

Probably my biggest learn in the last 12 months or so is the importance of time wrt user feedback.

Doesn't matter how good you think your code is written, or how well you think you designed it, or how well you think you tested it: Code isnt suitable for production until it has been hammered extensively by 10s of thousands of people in a process that takes decades (mostly because many of them get pissed at stuff not working/doing what they want, then check in a year or two later to see if what pissed them off is fixed - and only then actually complain if it wasnt and the project is still active)

X11 is well passed that - most notably with XCB.

Wayland hasnt even started that process yet, the lack of complaining when wayland was made default was interpreted as people being happy with the default rather than them just changing the default and quietly getting on with the day job.

(IMHO)