Linux vs Windows performance

[geronimo3]

Hello

i have searched this forum a little but found no discussion on this. i have installed a dualboot Linux Mint / Windows 10 configuration. both setups with the latest patchlevels and graphics drivers. i have expected to get some better performance out of the Linux installation, but i recognized the opposite. i have not executed a reproducable benchmark, but did some simple comparison flights with a new default start on both systems, Linux gave me approx 5% less fps than windows. I thought the unix kernel is running more efficient than windows 10, but the game performs slightly better under windows. maybe some lesser ram usage on linux.

are there any additional experiences in the comunity here ?

thx, geronimo

[fallenwizard]

Here on Arch with AMD & ACO shader compiler, the game runs pretty much on par with the Windows version. What's your graphics card?

[tomchk]

Do you typically get better performance in games on Linux? I never found that, as much as I hoped and was told to hope.

[fallenwizard]

Do you typically get better performance in games on Linux? I never found that, as much as I hoped and was told to hope.

It mostly depends on the graphics driver and if you are running a compositor. Those can seriously kill performance in games.

[geronimo3]

thanks for the comments, my graphics card is an geforce 1070. i must admit that i have no experience with Linux gaming. my assumption was purely based on studying the internet on the effectivness of windows and unix system cores.but yes i have no deeper knowledge. it was an interesting experiment anyhow and let me believe that windows10 is not a bad gaming platform. too me it looks that i´am not limited by the geforce 1070 rather than on the i7700hq cpu, because playing on the graphics setup do not show big improvments, but thats another discussion.


thx again, geronimo

[tomchk]

I think MS did quite a lot to improve performance in Windows 10 in general. It has worked on incredibly old and weak systems for me, and more smoothly than probably anything I've ever used. I hope it stays this way!

[Rastuasi]

We tried it on a friend's computer, as I'll never have a Windows system in my house, and he saw a 45% improvement on his machine when using my self modified Arch-based Linux kernel. You might need to make sure you're on the newest kernel and drivers (aka not Ubuntu or variants that don't have rolling releases).

[Angsaar]

We tried it on a friend's computer, as I'll never have a Windows system in my house, and he saw a 45% improvement on his machine when using my self modified Arch-based Linux kernel. You might need to make sure you're on the newest kernel and drivers (aka not Ubuntu or variants that don't have rolling releases).

If you wouldn't mind expanding on that, I'd be very interested in the .1% results of that experiment. The stuttery bits of X4 are the most annoying to me so if those see even a 10~20% improvement I'd be willing to split this SSD with something Manjaro-like (I'm mostly Debian-bound for work reasons and some stuff I play just doesn't seem to get along with it), average performance is fine at the moment.

[Panos]

We tried it on a friend's computer, as I'll never have a Windows system in my house, and he saw a 45% improvement on his machine when using my self modified Arch-based Linux kernel. You might need to make sure you're on the newest kernel and drivers (aka not Ubuntu or variants that don't have rolling releases).

Can you elaborate on this and your custom AMD kernel?

[Imperial Good]

A "45%" improvement in performance sounds like the original system was not functioning correctly. Kernel optimizations are meant to yield performance improvements of a few percent at most, especially in 2019 seeing how any such performance optimization has been critical since the OS first started to be used widely.