Cool that average FPS is better but:
The impressive FPS deltas aside, it should be mentioned that, with the exception of Arch Linux, average frame times (measured as 1% lows, in this case) on Linux were generally behind what Windows managed by up to 20%
I feel like worse 1% lows makes this title misleading. Hopefully with time this gap will close.
1 % lows are likely a driver thing (Nvidia calls it “Game Ready Drivers”), with Arch you’ll get new drivers (or kernel versions) much earlier, similar to Windows.
… Then why did they get the results that they did in the article?
The article specifies arch as an exception to that
Death, taxes and this post on Lemmy
i just saw this for the first time and it brought me joy
1 day cannot pass without this article getting reposted across various communities.
Yeah, these are getting really annoying.
I swear people just scroll through lemmy, see a post they like and then think to themselves, “this is cool, I should post this on lemmy!”
It’s okay. Lemmy isn’t a wiki. Content is organized temporally. Imagine these conversations as bar conversations (just because one group had a conversation one night, doesn’t mean another group can’t repeat it the next). If you are annoyed that the algo keeps giving you the same stuff, sort by All and New Comments and you’ll find niche communities to subscribe to.
As someone on Linux, and who thinks performance is generally slightly better on my machine after switching, I totally agree. This post has been old for a while now. Get some more data and then post that new thing or stop posting it.
A Lemmy option to hide posts of links already red in another post would be neat. (First time I see this one though)
Testing done on specific hardware and not a broad spectrum of machines is as relevant as asking one person their political opinion and saying that applies to their whole nation.
Well sure but rephrased it’s just “Three Linux distros that embarrass Windows 11 in gaming performance.” which to me, is equally interesting.
article title: windows DEAD LAST!
also in the same article: “… When it comes to FPS, the overall leader in testing was Nobara Linux, with Arch Linux and Pop!_OS trailing by 1–5%. Windows 11, however, was only 6% behind Nobara Linux. So, **there isn’t a massive performance delta here, **”
deleted by creator
is the point of article not to stroke the ego of the Linux absolutists that have some weird chip on their shoulders when it comes to video games?
deleted by creator
No it hasn’t, some games run better on the hardware tested when running Linux, some games don’t work at all on Linux whereas all games run on Windows.
Come back when they test multiple machines running various hardware and when they compare the experience setting up said machines to actually run the games.
You’re exactly the person this article was written for, someone who wants their opinion reinforced because they won’t take the time to analyze the data presented.
deleted by creator
gaming and the abundance of software and third party support and tutorials on windows is why I haven’t taken the dive to linux yet. So yeah, if linux does gaming as well or better my migration is more and more likely.
I’m gonna be that Linux bro and tell you to switch 😁 I made the full switch earlier this year and I’ve been amazed by how good proton is. There’s only been one game I couldn’t get to run until I did by installing some Microsoft VC runtime. Give it a shot! You just might be pleasantly surprised as I was.
deleted by creator
“on that one specific machine.”
You’re missing that part from your premise and it’s the important one.
Notice how they didn’t use one with an Nvidia GPU… Or even hardware released this year either…
Edit: Aaaw, I made you angwy and you downvoted me :(
The Nvidia drivers perform fine generally but it would be nice to verify it.
It’s funny because you say that and at the same time it’s the first question asked when someone has a hard time gaming on their Linux setup, “Nvidia GPU?”, even in this thread it’s come up.
The issue isn’t performance, it’s Nvidia’s unstable drivers.
E: fuck me, are people stupid? Performance and stability are not the same thing.
Performance on Nvidia cards on Linux is fine. The issue is the bizarre issues you have like multi-monitor weirdness or adaptive refresh rate not working properly. Nvidia’s drivers need kinks worked out but they aren’t slower.
Irrelevant to someone that wants their game to run.
I didn’t say otherwise. We were talking performance, not stability, that’s why I said the word performance, then said Nvidia’s drivers were unstable.
Understand? Performance means performance and stability means stability. I can appreciate that might be hard to grasp, but they’re different words for a reason, and that reason is they mean different things.
I don’t know why I bother talking to morons on Lemmy who deliberately misinterpret what people say and use that as a gotcha. You’re not smart for using a straw man argument.
Nvidia needs to sort their shit out.
You realise that not having access to stable drivers is a performance issue because it means games don’t run properly or at all?
“The issue isn’t my bike, it’s the bent wheels on it!”
Removed by mod
Nvidia isnt so bad if you’re on a stable distro it supports and using x(though Ive heard wayland support is improving for it). On rolling or more cutting edge distros where the kernel is likely to change every few weeks and major DE versions might ship that proprietary driver will hurt.
That said while amd is generally better on linux for this reason it’s worth mentioning that it has two huge flaws:
1.Its not perfect like the fans mention. As someone who owned a 3500u and 6650u apu life under amd isnt always sunny. 3500u had a kernel regression for about half a year that prevented the cpu from idling and rembrant apus have an issue where the whole system locks up which seems to come and go(feels like it’s gone for now but Ive thought that before). Desktop gpus are better, but they still did suffer from driver bugs. I think my experience with my 5600xt was better than windows fans had for that generation, but it was not entirely stable and I did suffer from many kernel panics and system freezes. A few mesa and kernel releases fixed that, but it wasnt perfectly smooth. In addition to that no hdmi 2.1 support which is fine unless you game using your nice oled tv because no tvs come with display port. Proprietary drivers do allow for supporting some of the more obnoxious features that arent allowed.
- It can vary gpu/cpu to gpu/cpu for how fresh your software will need to be, but generally newer hardware needs very new kernels just for basic support and it may need a few more releases to get stable or good. So if you want to just sit back with ubutnu LTS or debian you need to make sure the release cycle lines up with support for your hardware. The other end of the spectrum is that being on a bleeding or cutting edge distro can mean stability issues and regressions. So for example a month or three ago fedora pushed a kernel update that had a regression where my 6800xt gpu wouldnt clock up when utilized so gaming framerates tanked and retroarch shaders were choking up. I could just use the old kernel but I had to make sure that the kernel updates didnt bump it away. Also an entire point release and several releases after that before the bug was fixed.
So while there is a lot of pro amd comments in the linux world and its worth acknowledging that the open source drivers are generally good it’s not perfect and the grass isnt always greener.
Just from that comment we can see how far from mainstream adoption Linux is for gaming… You really need to want to understand how things work to fix things that might not work natively. Not every gamer wants to be super knowledgeable about computers, most just want to play games. Heck, I’m very good with computers and I know that what little time I have to play games I don’t want to spend trying to make them work…
What do the performance metrics look like for the games that won’t run on Linux?
About the same as Spiderman 2 or Ghost of Tsushima on Windows.
you mean the rootkits that won’t run on Linux?
When did ‘rootkit’ come to be a generic term for invasive software? Rootkits are a specific type of thing.
Vanguard, BattlEye, EasyAntiCheat, Ricochet, etc… all run in the Windows Kernel and most, if not all, have the functionality to run arbitrary code, so might as well class them as rootkits.
Anticheats that run in the NT kernel may as well be described as rootkits, especially as they aren’t transparent about exactly what they’re doing. Then there’s the question of what happens if they get compromised
If it has kernel level access and can run arbitrary code, that’s a rootkit.
It’s absolutely valid to call these systems rootkits.
Because “rootkit” sounds more ominous and scary than “kernel level anticheat” and the communities complaining about such things aren’t known to keep hyperbole to a minimum. Gotta push that FUD.
This article for instance, using language that insinuates a huge gap in performance between the Linux distros and windows, when it’s a 6% difference between the best and the worst, on one set of hardware.
deleted by creator
(They were not serious)
That’s the point they were trying to make. It was a facetious question.
I’ve been out of the industry for a while, but unless Windows was completely rewritten from the ground up in the last 5 years, this doesn’t surprise me. That OS has always been a hot, bloated mess. And no, I’m not a Linux bro. I use another heavily commercialised OS that doesn’t run Windows because I no longer have the energy to care.
An OS written on Unix can outperform Windows? I’m shocked.
windows is overbloated garbage
News flash, Windows is garbage.
Jawohl, jawohl.
I keep hearing and seeing from seemingly everyone that Linux gaming is better basically every month, how it keeps improving and stuff (like the article here)
But for me personally it never did in the last 5 years, whenever I try to step out if emulation and back to windows exlusove games? Its like 5 bullet Russian roulette, if it works at all and doesn’t stop working for inexlicitly no reason
What are yall doing to actually make things work somewhat reasonable (default lutris, proton, or ge has never even renowtly worked how well for me, at all)
Modern Linux kernel and steam with proton, and in a few instances lutris with wine. Unless it has anticheet, it’ll play pretty well.
what black magic are you using to make them work
AMD 7600x and 6700xt, Debian 12… proton, wine, keyboard and mouse? Been using it no problem with cyberpunk and Starfield for a few months now. Play Diablo 4 and overwatch with my kids. Been gaming on Linux for almost 4 years now. It HAS come a very very long way since the steamdeck was launched though. Proton and lutris are the glue that hold it all together.
idk what kinda glue you are thinking about, but it aint the one im seeing gtx 1660 super here and distro agnostic for me, same problems all around
cyberpunk, a slideshow at best, tried several times and several configs diablo 4 (got it from a friend), never launched, cause battlnet never works
I’m convinced it’s my AMD graphics that are making things so easy for me. I have had no issues at all with their drivers. Ran arch with no issues for a few years, now Debian for a few months. Have never had an nvidia card.
I have been using an all AMD system for years on Linux and haven’t had any issues. Some coworkers with Nvidia graphics said it was a nightmare. So it must be the AMD drivers
I have over 200 hours in cyberpunk on Linux. The gog version is a little bit more work to setup the steam version. If you have it on steam, and have steam installed natively (not inside wine) it should work assuming you have the correct GPU drivers installed.
I’ve always had weird, buggy shit with Nvidias Linux drivers. AMD is pretty great though.
You could try an open source game like xonotic that supports Linux to test as well.
For me default proton “just works” usually. But I play a lot of indie games
The only issues left with Windows-Only games is their crippeling-for-purpose anti-cheat code. Anything else works better on Linux.
So the question is whether to support those BDSM anti-cheat games, or get a better gaming experience.
the linux wars!! lets see who wins. also, its a friendly war, microsoft loves and will support linux
It’s probably more cost effective to run Azure Services on Linux machines than Windows machines.
They… accepted that they’re not killing Linux, yet. Back in the 2000s, before Azure, they felt quite threated by Linux and ‘Microsoft Linux’ was a joke. The then-CEO Steve Ballmer called Linux a ‘cancer’.
Microsoft has a how to for Installing Linux on the desktop.
Maybe. But it does so in ALL OF THE GAMES!!!
Unless you’re super into Fortnite or CoD, Linux plays virtually all of them too.
Linux has all the games I care about. So no need for windowd
I can’t get Spiderman 2 or Ghost of Tsushima to run on Windows. What am I doing wrong?
I’ve heard that Linux’s task scheduler is just much better than windows’, so it kinda makes sense that all would beat Windows.
A scientific judgement for sure.
I recently switched to Linux this year (finally), and my experience has been the same.
Not only that, but in some cases, playing a Windows version of a game with Proton seems to work better than the native Linux runtime.
Edit: I use Arch, btw. (lol jk I use EndeavorOS, which is based on Arch)
amen, i love EndeavorOS. i’ve jettisoned all Windows support in my house and anything that needs Windows gets put into an isolated VLAN that can’t talk to anything else. and for the archaic business crap that only has a Windows release, CrossOver is a godsend. same CodeWeavers devs that made Proton and is essentially Wine Premium.
I’m not an expert in networking stuff… If I am using a Windows 11 laptop (owned by my work) on the same network as my personal laptop while working at home, am I putting my privacy/data/etc. at risk? Should I be sequestering the work laptop in some way?
it wouldn’t hurt. i wish my work would just give me a VM to remote into instead of dealing with it on my network, at least in my case all the EDA tools I use are ran on Linux anyway… my last employer put so much
spyware“security” software on their work issued laptops that Suricata on my router/firewall would light up like a Christmas tree. no idea what it was trying to do without breaking out Wireshark and analyzing captures, but that’s when i said enough is enough… can’t be trusted.
This has been reposted enough to where I’m going to start blocking accounts that continue to post it 😎
I think it’s a combination of reposting on Lemmy, multiple communities posting similar stories, and news sites regurgitating results from other sites like it’s fresh news.
Agree. My gripe with this article is that I’ve seen it posted on ~6+ communities. I love that Linux is beating windows in gaming benchmarks, but I think the title sensationalizes it the out performance a slight bit.
Anyone have a good explanation on ‘Frame Time’? This is the first time I’ve heard of this term and after some quick googling I feel like I’m not understanding why it’s worth caring about.
I’m not surprised at the confusion, because they’re using it… not wrong, but very confusingly.
Frame time is literally the time to render a frame. So you’d expect that to be a number of miliseconds per frame and so for lower to be better.
But they’re not looking at frametimes, they’re looking at 1% lows and expressing that in fps, not in frametimes. So yeah, confusing.
For the record, the reson why the term is becoming popular is that there are now widespread visualizations that will give you a line of your frametimes in a graph so you can see if the line is flat or spiky. You’ve probably seen it on the Steam Deck or performance analysis videos or whatever. The idea is that all frametimes being consistent is better than high fps but low 1% or 0.1% low. So stable 60fps can look better than spiky 90fps and so on.
I interpret it as the time taken to render a frame. Unlike FPS which is basically a moving average (or rather 1 divided by the average frame time), frame time is a single data point. Collecting frame times allows you to do things like compute the median or, in this case, the lowest 1% of the frame times. That can give you a better idea of how smooth performance appears to the player, and what the worst-case performance is like.
