Benchmarking 9 Linux Distributions On A $50 Processor

[phoronix]

Phoronix: Benchmarking 9 Linux Distributions On A $50 Processor

Your choice of Linux distribution on a budget PC can mean the difference of ~14% performance overall. Here are benchmarks of Ubuntu, CentOS, Debian, EndeavourOS, Manjaro Linux, openSUSE Tumbleweed, Fedora Workstation, and Clear Linux on a $50 processor as we roll into 2020 with the newest Linux distribution releases.

http://www.phoronix.com/vr.php?view=28741

[sarfarazahmad]

Wow. Thanks for putting in so many distros. Fedora and Debian Testing look very promising

[danmcgrew]

DAMNED good article !

[muncrief]

Wow, if Clear Linux didn't have such abysmal package management I might consider moving from Manjaro. But it still looks like about the only real option is redundantly installing things over and over and over as mandated by separate flatpack "packages", and trying to deal with all the other isolation silliness inherent within.

[treba]

Nice to see Fedora being in such a strong standing. I just wonder again whether the Firefox numbers have to do with the fact that they are the only ones using the Wayland backend or that they are using LTO/PGO (https://mastransky.wordpress.com/201...-with-pgo-lto/). Anyhow, the probably most active Firefox on Linux dev, Martin Stranky, is the Fedora Firefox maintainer, so they should be in a good standing here.

[S.Pam]

Really good article!

Mind posting the dmesg output from Ubuntu or any of the other ones? I never got any distro to work on my 3000G APU without stack traces/oopses. Look here for example: https://www.phoronix.com/forums/foru...recent-kernels

[perpetually high]

Thanks Michael, this is good stuff.

As an Ubuntu user, every time I see Clear Linux destroy Ubuntu and other distros it makes me feel better about all the damn tinkering I do with compiling the kernel, out-of-tree patches, OS settings, etc. That stuff makes a difference! (insert "C FLAGS MATTER!").

I'm almost certain if I ran those benchmarks on my machine I'd fall in line closer with Clear Linux than Ubuntu 19.10. I might be totally wrong and I should just run the benchmarks instead of heresay, but that's my feeling. At the end of the day, we're all using Linux here, and Clear Linux is proof of a distro that tweaks the crap out of everything to squeeze out performance, and the results show.

[DanL]

You should run update-pciids on Ubuntu 18.04.x, Debian, etc. before benchmarking to make the info table more readable.

[ms178]

Thanks Michael, this is good stuff.

As an Ubuntu user, every time I see Clear Linux destroy Ubuntu and other distros it makes me feel better about all the damn tinkering I do with compiling the kernel, out-of-tree patches, OS settings, etc. That stuff makes a difference! (insert "C FLAGS MATTER!").

I'm almost certain if I ran those benchmarks on my machine I'd fall in line closer with Clear Linux than Ubuntu 19.10. I might be totally wrong and I should just run the benchmarks instead of heresay, but that's my feeling. At the end of the day, we're all using Linux here, and Clear Linux is proof of a distro that tweaks the crap out of everything to squeeze out performance, and the results show.

Aye! Inspired by the recent scheduler discussions, my own experiments with a tuned kernel were equally positive. I did get a much better gaming experience with a kernel which was compiled with aggressive flags and tuned for low-latency - I also tried out the new BMQ scheduler which worked really well.

[Random_Jerk]

Manjaro is not good or bad at anything. It sits dab right at the middle of the pack in almost all tests