just benchmarked a game on both windows11 and linux and thought i'd share
win11 fresh install with core isolation off

frame gen on?

linux cachyos fresh install with mitigations=off in kernel, proton-cachyos for proton and game-performance for power profile

>>716490807
settings ultra, frame gen off, dlss off

>>716490695 (OP)
>>716490995
Results aren't too surprising for an Nvidia card.
Funny that the driver is 2 major versions behind.

>11
why?

>>716490695 (OP)
the wilds benchmark tools mix in cutscenes to bias the result
you should do your test with raw gameplay

here's proton experimental without power profile set
also there's visual artifacts in some parts (when the main character is riding the bird through the desert)
>>716491141
i dont think there's a newer version of the nvidia driver for my distro at least

>>716491327
forgot pic

>>716491215
why not

>>716491220
in the section where the main character is riding the bird through the desert (gameplay section) i notice the fps was around 10fps lower for linux
70~80fps vs 60~70fps

>muh fps
Has anyone ever chosen between Windows and Linux based on benchmarks? I feel like there's no reason to even think about smallish frame-rate differences unless you're fully comfortable with both operating systems and enjoy using them. Most people would either hate using Linux or are using Linux because they hate Windows. Nobody should constantly put up with an OS they don't like to use on a daily basis just for +10 fps. Unless the performance difference is enormous like 60fps vs 15fps (and it rarely is), benchmarks are just something to wank about after you've already made your choice.

>>716492602
I chose linux because I hate windows and unironically love systemd.

>>716492602
i do, guess why i have core isolation off and mitigations=off

im running a few more tests
here's nvidia closed source driver

splitlock mitigate off
nvidia closed
nvidia gsp disabled

>nvidia
yep

this thread would be very informative if monhun wild's benchmark wasn't shit.
80% of it doesn't put the test through normal gameplay situations, only hovering around idle towns and pre-calculated cutscenes.

>>716490695 (OP)
Is this Character Action Benchmarking?

>>716490695 (OP)
Did you properly lock the GPU and CPU to maximum speed?

>>716495690
Yep. But only the benchmarking has arcade scoring.
What a garbage ass series for brown people monster hunter is

>>716495482
cant think of another good one to use that's not western AAA shit
there's gameplay-ish sections and i can clearly see the FPS difference there though
>>716495690
it's the one you click download on the main game store page
>>716495827
all of the tests besides one here>>716491546 are run with game-performance power profile as instructed here
https://wiki.cachyos.org/configuration/gaming/#power-profile-switching-on-demand

>>716490695 (OP)
I've read cachyos gives lower fps compared to other distros because it's settings are made to prioritize latency instead of average fps. Is that true?

>>716496060
>cant think of another good one to use that's not western AAA shit
at that point you should stick to actual benchmark tools.
at least you're comparing different setups with the same benchmark, however you will get a lot of run to run difference, and I have no idea how bad that is with this benchmark.

>>716496227
i picked cachyos exactly because i saw it on youtube that it is consistently the best performing distro in terms of FPS along with bazzite
didnt pick bazzite because i have some linux experience and bazzite looked more like a distro for noobs

>>716490995
loonix can't count to 32

>>716490695 (OP)
>>716490995
>>716491546
>>716494565
>>716495315
i did some benches like this when it came out to get a feel for win10 vs gentoo
will redo the test with a RTX PRO 6000 Blackwell when driver 580 comes out

if anybody wants me to do additional tests just post here before i fuck the fresh installs up with bloatwares

>>716497595
you got a windows 10?

>>716498251
no

>>716497120
use case for counting to 32?

>>716490695 (OP)
What is core isolation and how do you turn it off?

>>716490695 (OP)
>>716490995
>>716491546
>>716494565
>>716495315
You should be getting 2x those FPS with how average Wilds looks

>>716500938
1 second in your favourite search engine
disabling it boosts your pc performance (more than 10% extra fps in most cases) at a cost of theorical increased risk of getting hacked in which nobody has ever seen demonstrated outside the lab

>old as fuck cpu and gpu
>WAAAHHHH WHY ISN'T IT 200FPS ON MY TOASTER
The game runs fine and looks great, stop being poor.

>>716490995
>mitigations=off
That's LESS performant on modern CPUs. Your test is invalid.

>>716502145
I have a PC with RTX 3080 but its a laptop. How well does a Desktop RTX 3080 run Wilds? Anyone?

>>716490995
>mitigations off on a am5 part
I thought 7000+ ran worse with them disabled.

>>716496227
Lower frame latency = more completed frames in a given time frame = higher FPS

>>716490695 (OP)
Core Isolation is a meme now, you need to disable Virtualization Protection and Windows Hello completely.
Then use the Core Parking app to unpark all cores.
now the game should run a lot better.
Also the benchmark is flawed because it doesn't have the anti tamper crap they added in the main game, so it's meaningless.

>>716502145
this, just wait a 6090

>>716490695 (OP)
>33
Nice masonic numbers faggot

>>716506525

Guide for all of this?