As I watched the results, having made similar tests myself, I was in awe at how much variation you found. But then it hit me, your numbers are consistent with Ryzen pre-patch, check your video description and there it is, Ryzen as I suspected. You probably weren't fully updated on Windows 11.
Would have liked to see a comparison of DirectX on Windows and Vulkan on Linux (or even better Vulkan on both) - OpenGL just sucks in general and it doesn't really matter to me which platform does slightly worse in it. Vulkan and DirectX however are both great APIs and would be a better performance comparison. Other than that, great overview and testing :)
@@etaashmathamsetty7399 Sure, but it's not anymore and in pretty much every benchmark OpenGL is far worse than Vulkan, so I don't see the point. It's like saying "oh in Linux we used to only have the terminal and no GUI so why would anyone do benchmarks with GUIs". I'm not saying that he shouldn't have done any OpenGL benchmarks, just that I'd have liked to *also* see Vulkan benchmarks.
@@chair-mander I wasn't really trying to prove anything and I agree with you, but vulkan still isin't super widely used and devs prefer to use directx. (also there's less hardware support for vulkan) I do think that vulkan is better, but those idiots are still to stubborn to switch to the more versatile vulkan. I think they feel that if they add vulkan, they have to add opengl too.
I actually got a Thinkpad problem and therefore I own laptop from t61 to t480s. I tried win10 vs Linux on every one of those and I can say: no you don't, if you put 8gb of RAM in them and use Gnome which is the defacto workstation choice. Using a tiling WM of course it feels snappier a bit beating out Windows and osx (i got a Mac too and acess to some more. Osx got the most slowed down with the last decade).
@@Hardcore_Remixer Xerife Tech is that you?hahaha,his isos helped me a lot with Virtual Machines,when i run some Windows Isolated App,inside my Linux Workflow.....
One of the speed advantages of Linux that you didn't test was shutdown time. It's weirdly slow on Windows, and Mint feels like 5x faster. It doesn't matter if you're just shutting your computer down for the night, but it's really nice if you want to reboot
When you "shut down" windows, it creates a hibernation file to make the next startup faster.. Its not a apples/apples comparison. you can disable fast startup if you want shutdown to run faster.
@@Blinknone isn't hibernate supposed to be for that and shutdown just do the regular shutdown? I could never figure that one out, you need special kind of shutdown to make a startup faster?
Neat to see, though using Flatpak basically nullifies testing advantages and disadvantages of each distro's core libs and platform files. Also, even though I don't use it, it's weird not seeing Clear or Solus here, it's the posterchild of the linux drag race.
As a dual-booting Mint user who mainly uses his PC for web-browsing (Mint) and gaming (Win 10): These results look good. Completely by chance picked the right Distro for the job :D Also: It is pretty noob-friendly. So that's definitely also a plus.
Same here! I used to use Windows before, but when I just switched to Linux, I started to really hate nearly all Microsoft for well-known reasons. My first ever use of Linux Mint was the biggest relief in my life.
I wish I could use Mint but it shuts itself every time I try to use the latest nvidia driver with the latest kernel version 😢 been a while since I tried, maybe I should try again
It worth mentioning that these benchmarks were run on an rx580 (hey thats my card!) and the reason that OpenGL apps ran so bad compared to linux on windows is AMDs fault for having a terrible OpenGL implementation in their windows drivers.
@Mike Valadez actually, he mentioned that opengl is probably a disadvantage for win11. the cs:go, blender render and geekbench are all good and representative tests.
@Mike Valadez Yeah! What a guy! He cherry picked OpenGL scores to make Windows look bad... and then proceeded to show DirectX performance to make it look good? I mean, it was pretty impartial, there were no biases here.
I've been thinking about switching from Linux Mint to Fedora, and this shows some very interesting information. I like Mint and how fast it is, but the UI on Fedora seems to be more user friendly without giving up too much speed.
I needed to change because my dell G15 laptop dont accepted another screen with hdmi. Fedora just worked... But what I do is have a major system and use a virtual machine, so if I do mistakes I just delete it...
I mean, it’s solid, but I’d say that the big takeaway is that Linux is solid all around. Mint lead in some things, fedora in some thing and Manjaro in some things. Given the nature of desktop Linux, I don’t think enough is changed “under the hood” that distro really matters. I use fedora on my home media server and Manjaro on my audio station/coding/gaming rig. Only thing I have against Mint is that it is Ubuntu based. And even then, there are worse things than Ubuntu.
Linux Mint is great and really addictive. I just don't like to game on it. The graphics don't look as good as compared to other distro's or maybe Windows but in General the Graphics appear slightly darker and it looks and feels different not in a bad way and it will certainly not hamper your gaming experience but still. I use a GT 1030 ddr5. If anyone knows how can we fix it then it would be awesome
And this is why i love Linux in general, especially PeppermintOS. Using Windows 10 on a laptop with a 3rd gen Intel CPU and HDD is absolutely S L O W, but with PeppermintOS it's actually quite tolerable. But i betcha using Both Windows 10 or Peppermint with an SSD will drastically improve both boot times. (Not sure who would win here) But I'm sure PeppermintOS would knock it out of the park when using an SSD. (And if you still occasionally need Windows, you can always make a Dual-boot setup. Install Windows first, then Linux)
Try the Garuda distro it's becoming just as Legendary as Windows takes a few moments to install controller drivers but works a hell of a lot better than SteamOS it also offers emulator downloads through VIA Add Software module and comes equipped with Citra if you want a modern Linux it's definitely worth looking into if you still exist I hope you take my suggestion.
An idea for a new video: Desktop environment vs. Window manager What is the difference. Commons projects. Can I swap as I please or does one require a specific other (eg. GNOME requires Mutter).
The only things that could really be tested with raw Window managers is the screen compositor and if it uses wayland. Picom, and the multi abundance of forks out there, is the most widely used X screen compositor.. Other than Compiz for Xfce and the dieing Compton. - Picom is an active fork of Compton Wayland I guess would be pretty broad. It will depend on how they utilize Wayland in their window manager.
Switching the Desktop Environment's window manager is kinda hacky and might end up with weird results.. Few examples: - Gnome would either need a legacy build or you would need to do some hacking. Try looking at the i3-gnome project and/or sway-gnome project. - Plasma's disconnect is really apparent when you see the WM launch and fully working.. But you got to wait for Plasma's bootup screen to end before you can use the desktop. - Xfce will have limited customizing support. It would have to be customized through the WM you choose. Basically you'll have to bypass the DE's WM in certain ways to switch the WM. Customizing window borders and title bars will depend on the WM you choose. End results may vary. What you are really getting out of this is the DE's compositor and it's status bar.
@4:20 , AMD drivers on windows doesn't have proper support to OpenGL anymore, that's why. score would be the same as others if that wasn't the case. Edit: oh, you can see that in all of the tests after that are using openGL.
Were these tests ran before or after the patches for AMD CPUs on Windows? Seemed like cpu stuff lagged behind and anything with opengl lagged behind in Windows because of AMD's OpenGL implementation on Windows.
Huh, is there no DXVK or Vulkan-based graphics benchmark tools? While Windows with Direct X is usually at the top, Linux with Vulkan/DXVK usually comes in a punching distance from the comparisons I've seen (though those are oriented to gaming). And I wonder what exactly causes all the differences between each systems though - I guess it's whatever services and process they have on the background (which includes the DE but I wonder if libraries can causes variance) as well as how they build their kernels? Very interesting benchmark nonetheless, though. Been using Win11 in a VM, and so far it seems to go well as a VM OS over Win10. If I ever install it as a secondary OS, it'll be on the day when I get off my ass to actually do an Arch + Win11 install.
Yeah, I was wondering the same. It usually perfoms better native Vulkan builts on Linux rather than Windows anyway. It also heavly depends on the game too.
do it man, I have my Arch + Win 11 on my laptop for a month now and it's great. Do everything on Arch and only have to hop to Windows for very specific task that I can't do on Arch
Thanks for this. It would have been good to include windows 10 test as well as the people who are thinking of upgrading will get to know the difference
Hi, Techy, could you do a video on distros recommended to use with NVIDIA cards? I'm having a hard time with a rtx 2060 laptop on Fedora 34 with Wayland, whenever I plugin an HDMI cable the whole system freezes, the only way out is the power button :/
Liar boot test, since a unique Windows 11 feature was not used. The fast boot. My Windows 11 here starts in 5 seconds after turning on the machine. There's something wrong with this test. Not to mention, that count for Windows 11, started even before the computer was turned on. Only at 6 seconds, entered the screen where the BIOS of the motherboard is operated and not the OS. While the Linux distros, they were already on the system loading screen. This is clearly a manipulation to benefit your personal taste.
There's a difference in speed between Manjaro and EndevourOS? And if DE's can impact in graphical performance, for example GNOME vs LXDE or something like that.
What file system did Mint and Manjaro use for the bench marking? I wonder if the file system also affects the file read/write and even execution speed.
I would love to see a deep dive into what plays into these performance differences. For example if you profiled some of these applications to see which calls take longer, what versions of libraries they call into, how much time syscalls contribute, etc.
if virtualization is ON on the BIOS a fresh install of windows 11 will activate Core Insolation, that virtualizes the kernel for security, that has a porcentual performance lost, be sure to deactivate it if that's the case (since virtualization is not active in most systems by default)
LOL. ( facepalm) It doesn't matter if it's virtualized or not....Windows has PROVEN for 3 decades straight to NOT be a secure Operating system. changing the icy of the cake still means..it's a cake Now I agree with your point that this virtualization being on in bios CAN hinder performance in gaming and a few things by a small but noticeable enough margin. That's also been proven since winturd 11's launching.
Glad I bit the bullet and went down the Mint & Win10 dual boot route a few weeks ago. I'd been on Win8.1 since 2015 but I was increasingly encountering software that had warnings about Win8.1 incompatibility. I'd installed Elementary OS on an old laptop about 5 years ago just to get acclimatised to Linux before letting it anywhere near my desktop, so the learning curve for Mint wasn't too steep. Win 10 stripped of all telemetry, bloatware, and similar nonsense is actually quite beautiful, but to maintain it in that pristine state is too much for me. I want to get things done, not tinker with the OS constantly.
@cross_talker Yes, it's completely safe. Just be sure you write to right partitions/drives etc. I'm more worried about linux side, lol. I have read that windows tends to overwrite grub bootloader with its own thus making you unable to boot into mint. Haven't tried dualbooting myself but it should be pretty easy to reinstall grub from any linux usb.
you had a fresh windows 11 install right? at the default there is a feature called VBS security enabled which really make the windows 11 graphical performance worse.
I've done scientific/engineering simulation benchmarks in linux and I've seen fedora peak way higher than arch, so manjaro should be same. All ubuntu derivatives fell behind.
Please do this with windows 11 insider beta they have done insane performance improvements! up to 2x faster startup 4x faster shutdown 2x faster opening apps. This new windows update is insane! Edit: And remove most telemetry! Would like to see the difference then! For now im keeping notifications on and not switching to a linux distro Edit V2: I switched to linux anyway 5 months later.
You could test Nobara and different browsers with every system to display which one is the fastest option. The tests felt simple and not really conclusive. Which kernels were the operating systems on? Which version were you using? etc.
Good stuff all around, except for testing Kdenlive Nightly on Windows, and choosing Edge (which just barely came out of beta on Linux) for the browser perfomance comparison; both of these favor W11, especially in regards to Edge running on Windows. Also, no ram usage comparisons?
The Geekbench for WIndows was done with the 'Balanced' power plan, I wonder how much better Windows would have been overall with the performance power plan
Arch with KDE can't run smooth live stream in browser + discord while watching game stream + installing some soft. When I run new mid CPU usage task, sound start stuttering while this task loading. Yes. it fast, but looks like not good resource using. I mean, already running task cannot get enough resources for smooth working for a while. My win10 start discord little longer, or something else, but already working tasks still works as must, until I not start something that use 100% CPU time.
Can you make a tutorial at how to install Davinci resolve and amd opencl on manjaro. I've been try using some package in AUR, and still can't using Davinci resolve
I've been rocking Fedora 35 on Wayland with the Nvidia 495 driver and it's great and buttery smooth! Just make you get the dash to panel extension to make it similar to Windows. People are sleeping on this big time. Actually no, it's literally somewhat of a new thing lmao.
@@outofahat9363 as if right now, the implementation on fedora is the only official and native implementation of nvidia on wayland, definitely won't be 100% stable, but from what I heard, it's pretty good already
Why my linux mint boot slow about 33 sec, my laptop specs is amd a8 6410 , 8 gb ram, 128 gb ssd. But if i using windows 8.1 it boot really fast just about 5 sec
Kudos for this! This is one of the few fair "Linux vs Windows" comparisons I've seen on YT. So many of them are so blatantly biased one way or the other, and it's so disingenuous and unhelpful. So, again... Kudos for keeping it as fair as possible.
I noticed such a thing. I have Internet 100 mb. on Fedor, the speed test shows 55 and on Manjaro 75. and Fedor in live mode shows 75, but after installing on disk 55. can you do the test
AMD OpenGL performance on windows is bad. That’s why when using directX it was way ahead of Linux. Hopefully AMD address it sometime this century. Vulkan vs DirectX would be good to see
What the hell? My windows 10 installation boots way faster than my linuxmint install on the same machine. How did you get yours to boot in just 13 secs?!
As a huge Microsoft fan, it disappoints me how Microsoft is not doing as well as they did back then, Windows 10 was the best, Windows 11 has serious problems, I'm very impressed with Linux and the way they each performed was really cool to watch
hey techHut can you make a guide on what to do if fedora has very bad video playback? i love fedora but when i used it playing video was just so horrible. i tried the ffmpeg fix the forums said but even that didn't help not that i am totally taking the L i went on to Manjaro. but i rly wanna give Fedora a chance
add few gaming benchmarks like civilization build in turn time benchmark. it feels to me like your blender isnt configured to use hardware acceleration properly
Remember that GIMP is based on the GTK framework and is not easily compatible to Windows and has to go through a built-in compatibility layer of some kind.
Comparing games by average FPS is really stupid way to do it. 1% lows is the important value since that is the most distracting aspect, the FPS dips and frame times is what we are sensitive to.
Yes..but that DOESN'T show how often within any individual game that it DIPS to that lowest/worst case fps......THAT is the most critical deal followed by what fps the game MOST OFTEN runs at.
Why do you use Edge instead of Chrome in web benchmark? I believe Edge is unoptimized in Linux. I tested Chrome base in Linux that Edge has the lowest score compare to Chrome base such as Chromium or Chrome.
the only reason windows was faster in your kdenlive test is probably because you used flatpak. A native package will definitely squeeze out more performance.
I'm done. If you don't understand why Windows would beat out flatpaks, you haven't used Linux enough. KDENLive is available (native) in pretty much all distros. It's just weird that you would use what essentially a Java KVM to run against Windows native.
Needed to open the same exact project requiring the same exact version on 4 different platforms. Flatpaks are the direction we're going in general so I figured it was a fair test. Also watch out for the video I make on this topic. Rendering doesn't give you that much of a difference compared to installing directly from the repo.
@@TechHut I will be honest and say that USING programs in flatpak form is not something I have a lot of experiene with...so, I am looking forward to the rendering. But, since flatpaks are the basic equilavent of running programs in a virtual environment, I stand by my objection using them to determine load times... cuz, quite obviously, the virtual environment has to load first. If it works better (or even comparable) inside that environment, fine! I will accept that.
These benchmarks are to taken with a grain of salt,, Different kernels, latency settings in the linux distro's for a start, The file copy test is out, as linux writes to ram first then disc.
A lot of these were really hardware benchmarks, like grinding on the CPU to encode a video, the OS doesn't affect that much. I'd like to see some benchmarks that cover the human usability stuff, like in Windows if you open your downloads folder with like ~100 files in it, it takes multiple seconds for the directory contents to appear, making the OS feel very slow in day to day use. Also the amount of time it takes for the right click menu to appear. The Windows start menu is also horribly slow, often showing blank UI for a few frames before text labels appear. Launch times for various programs would be more useful than boot time (since I rarely reboot), especially when you have a few programs running already.
Hi, I have an old dual-core pc higher than 2ghz and 2 giga ram and a minimum gpu I'm interested in the best programs for photo and video editing which distro do you recommend for more versatility? Thank you.
Set up your own Linux server with LINODE : www.linode.com/techhut
$100 - 60 day credit on new Linode accounts for TechHut viewers.
thank you so much
I hope you can add an Arabic translation
I'm following you, but we don't know English. Thank you in advance, sir
As I watched the results, having made similar tests myself, I was in awe at how much variation you found. But then it hit me, your numbers are consistent with Ryzen pre-patch, check your video description and there it is, Ryzen as I suspected. You probably weren't fully updated on Windows 11.
bro you looks like Magomed Ankalaev, russian ufc fighter
The fact that all the Linux distros are free and don't require you to upgrade your PC or sign up to use them is the big win.
Yes for a home-user. If you're looking for a computer job - it's best to know Windows and Linux... and how to program on both.
@@FeedScrn Who don't know how to use Windows?
@@unknown_codec_404 - Agreed.
@@gomes-fonseca disagree. FOSS is a clear exception to that rule because all tracking is public.
@@FeedScrn People who know how to use Linux will do just fine with Windows.
Now give us a AmogOS benchmark comparison
Now that is a sus comparison
Wholesome keanu chungus approves this message
openSUSE vs AmogOS
in a aSus computer
Pretty sus
Imagine how much faster Windows would be without all that spyware
what spyware?
@@orkhepaj u must be a windows/ms user
@@affyne it is the best os so far
The thing slowing it down in the Graphics tests us always OpenGL
@@orkhepaj For spying, undermining the user's control and building up a dependency cage etc.?
Well then for sure it is one of the best OSes.
Would have liked to see a comparison of DirectX on Windows and Vulkan on Linux (or even better Vulkan on both) - OpenGL just sucks in general and it doesn't really matter to me which platform does slightly worse in it. Vulkan and DirectX however are both great APIs and would be a better performance comparison. Other than that, great overview and testing :)
Opengl used to the only thing u had on Linux
@@etaashmathamsetty7399 Sure, but it's not anymore and in pretty much every benchmark OpenGL is far worse than Vulkan, so I don't see the point. It's like saying "oh in Linux we used to only have the terminal and no GUI so why would anyone do benchmarks with GUIs".
I'm not saying that he shouldn't have done any OpenGL benchmarks, just that I'd have liked to *also* see Vulkan benchmarks.
@@chair-mander I wasn't really trying to prove anything and I agree with you, but vulkan still isin't super widely used and devs prefer to use directx. (also there's less hardware support for vulkan) I do think that vulkan is better, but those idiots are still to stubborn to switch to the more versatile vulkan. I think they feel that if they add vulkan, they have to add opengl too.
@@etaashmathamsetty7399 Ah I see! My bad for misreading then :) For DirectX apps though Vulkan performance is still great because of DXVK
now this zink for open gl on linux works similar to what dxvk does with vulkan
You will really see the difference in preformance when you do this bench mark on a low -mid end system.
I actually got a Thinkpad problem and therefore I own laptop from t61 to t480s. I tried win10 vs Linux on every one of those and I can say: no you don't, if you put 8gb of RAM in them and use Gnome which is the defacto workstation choice. Using a tiling WM of course it feels snappier a bit beating out Windows and osx (i got a Mac too and acess to some more. Osx got the most slowed down with the last decade).
Yeah, but compare Stripped versions of Linux like Mint xfce with Windows 10 LTSC, Superlite or Ghost Specter if you wanna make it fair.
@@Hardcore_Remixer Xerife Tech is that you?hahaha,his isos helped me a lot with Virtual Machines,when i run some Windows Isolated App,inside my Linux Workflow.....
Fedora is fast mostly because how vanilla the packages and the kernel, and how well mutter, wayland and flatpak is integrated into it.
Well using Flatpak version could have been an issue..
One of the speed advantages of Linux that you didn't test was shutdown time. It's weirdly slow on Windows, and Mint feels like 5x faster. It doesn't matter if you're just shutting your computer down for the night, but it's really nice if you want to reboot
When you "shut down" windows, it creates a hibernation file to make the next startup faster.. Its not a apples/apples comparison. you can disable fast startup if you want shutdown to run faster.
@@Blinknone I didn't know that, interesting. I assume disabling it would make the next startup slower, so it'll be bad for reboot time either way.
@@NovemberOrWhatever hahaha😂
@@Blinknone isn't hibernate supposed to be for that and shutdown just do the regular shutdown? I could never figure that one out, you need special kind of shutdown to make a startup faster?
@@fulconandroadcone9488 they changed the shutdown, now it does the same thing hibernate does
1:44 i love your hand movement when showing us the USB!
Neat to see, though using Flatpak basically nullifies testing advantages and disadvantages of each distro's core libs and platform files. Also, even though I don't use it, it's weird not seeing Clear or Solus here, it's the posterchild of the linux drag race.
0:48 lol this is the clip of my Dell R710 server I recorded and published on pexels. Glad you liked and used it 😅
As a dual-booting Mint user who mainly uses his PC for web-browsing (Mint) and gaming (Win 10): These results look good. Completely by chance picked the right Distro for the job :D
Also: It is pretty noob-friendly. So that's definitely also a plus.
@@mr.mastermind4840 Yeah, it may get stuck when you play a modded game 🤪
@@hahaitsme found bill gates account
Same here! I used to use Windows before, but when I just switched to Linux, I started to really hate nearly all Microsoft for well-known reasons. My first ever use of Linux Mint was the biggest relief in my life.
I wish I could use Mint but it shuts itself every time I try to use the latest nvidia driver with the latest kernel version 😢 been a while since I tried, maybe I should try again
Don't you have to restart your computer every time you want to switch tasks?
I'm actually impressed by Mint's performance.
Question: Did you disable the spectre/meltdown mitigations on your Linux boxes?
tengo ordenador no tengo
eso no veo entonces no quiero tener
veo que ya no veo no sé que no veo y no puedo tener ordenador porque no tengo
@@catalinaleon8040 what the...
@@catalinaleon8040 that sentence doesnt make sense.... ni siquiera en español
It worth mentioning that these benchmarks were run on an rx580 (hey thats my card!) and the reason that OpenGL apps ran so bad compared to linux on windows is AMDs fault for having a terrible OpenGL implementation in their windows drivers.
We just don't use their implementations on linux
@Mike Valadez actually, he mentioned that opengl is probably a disadvantage for win11. the cs:go, blender render and geekbench are all good and representative tests.
@Mike Valadez Yeah! What a guy! He cherry picked OpenGL scores to make Windows look bad... and then proceeded to show DirectX performance to make it look good? I mean, it was pretty impartial, there were no biases here.
I've been thinking about switching from Linux Mint to Fedora, and this shows some very interesting information. I like Mint and how fast it is, but the UI on Fedora seems to be more user friendly without giving up too much speed.
I needed to change because my dell G15 laptop dont accepted another screen with hdmi. Fedora just worked... But what I do is have a major system and use a virtual machine, so if I do mistakes I just delete it...
How about features in mint compare to Ubuntu and manjaro???
or tumbleweed
I'd suggest Linux Mint over everything, because of its usability and accesibility and thanks to it's base (debian through ubuntu) it's rock solid.
@@DucLuuGmail cinnamon
Make a speed test between OPENSUSE TUMBLEWEED vs FEDORA 35 Gnome it would be really interesting...
This shows that Linux mint is an extremely great choice
I want to build new PC in the future using Linux Mint. Everything I read about them looks great and accessible for beginner
besides, it is easy to use for window users, I quickly got used to it, but left the windows as a second OS
I mean, it’s solid, but I’d say that the big takeaway is that Linux is solid all around. Mint lead in some things, fedora in some thing and Manjaro in some things.
Given the nature of desktop Linux, I don’t think enough is changed “under the hood” that distro really matters.
I use fedora on my home media server and Manjaro on my audio station/coding/gaming rig.
Only thing I have against Mint is that it is Ubuntu based. And even then, there are worse things than Ubuntu.
@@dstinnettmusic you might want to check out LMDE (Linux Mint Debian Edition)
Linux Mint is great and really addictive. I just don't like to game on it. The graphics don't look as good as compared to other distro's or maybe Windows but in General the Graphics appear slightly darker and it looks and feels different not in a bad way and it will certainly not hamper your gaming experience but still. I use a GT 1030 ddr5. If anyone knows how can we fix it then it would be awesome
i think these benchmarks will be different when you use intel processors
You did the best thing you could there, you suggested a direction for future study.
And this is why i love Linux in general, especially PeppermintOS.
Using Windows 10 on a laptop with a 3rd gen Intel CPU and HDD is absolutely S L O W, but with PeppermintOS it's actually quite tolerable.
But i betcha using Both Windows 10 or Peppermint with an SSD will drastically improve both boot times. (Not sure who would win here)
But I'm sure PeppermintOS would knock it out of the park when using an SSD. (And if you still occasionally need Windows, you can always make a Dual-boot setup. Install Windows first, then Linux)
Try the Garuda distro it's becoming just as Legendary as Windows takes a few moments to install controller drivers but works a hell of a lot better than SteamOS it also offers emulator downloads through VIA Add Software module and comes equipped with Citra if you want a modern Linux it's definitely worth looking into if you still exist I hope you take my suggestion.
An idea for a new video: Desktop environment vs. Window manager
What is the difference. Commons projects. Can I swap as I please or does one require a specific other (eg. GNOME requires Mutter).
The only things that could really be tested with raw Window managers is the screen compositor and if it uses wayland.
Picom, and the multi abundance of forks out there, is the most widely used X screen compositor.. Other than Compiz for Xfce and the dieing Compton. - Picom is an active fork of Compton
Wayland I guess would be pretty broad. It will depend on how they utilize Wayland in their window manager.
Switching the Desktop Environment's window manager is kinda hacky and might end up with weird results.. Few examples:
- Gnome would either need a legacy build or you would need to do some hacking. Try looking at the i3-gnome project and/or sway-gnome project.
- Plasma's disconnect is really apparent when you see the WM launch and fully working.. But you got to wait for Plasma's bootup screen to end before you can use the desktop.
- Xfce will have limited customizing support. It would have to be customized through the WM you choose.
Basically you'll have to bypass the DE's WM in certain ways to switch the WM. Customizing window borders and title bars will depend on the WM you choose. End results may vary.
What you are really getting out of this is the DE's compositor and it's status bar.
@@Ironpants57 you can disable plasma's launch screen.
@@AbteilungsleiterinBeiAntifaEV b-but the launch screen is important
Flatpak removes most of the differences, basically like running it from a container. I'd be curious to see what it would run like without that.
@4:20 , AMD drivers on windows doesn't have proper support to OpenGL anymore, that's why. score would be the same as others if that wasn't the case.
Edit: oh, you can see that in all of the tests after that are using openGL.
Were these tests ran before or after the patches for AMD CPUs on Windows? Seemed like cpu stuff lagged behind and anything with opengl lagged behind in Windows because of AMD's OpenGL implementation on Windows.
After. I wanted do this video when Windows 11 first dropped, but I had to wait for everything to get fixed.
@@TechHut Very good to know and makes this info much more accurate. Windows has always been slower than Linux in cpu related tasks but wow.
Would you be able to tell a speed difference without benchmarks? Probably not, but still interesting.
Huh, is there no DXVK or Vulkan-based graphics benchmark tools? While Windows with Direct X is usually at the top, Linux with Vulkan/DXVK usually comes in a punching distance from the comparisons I've seen (though those are oriented to gaming). And I wonder what exactly causes all the differences between each systems though - I guess it's whatever services and process they have on the background (which includes the DE but I wonder if libraries can causes variance) as well as how they build their kernels?
Very interesting benchmark nonetheless, though. Been using Win11 in a VM, and so far it seems to go well as a VM OS over Win10. If I ever install it as a secondary OS, it'll be on the day when I get off my ass to actually do an Arch + Win11 install.
Yeah, I was wondering the same. It usually perfoms better native Vulkan builts on Linux rather than Windows anyway.
It also heavly depends on the game too.
do it man, I have my Arch + Win 11 on my laptop for a month now and it's great. Do everything on Arch and only have to hop to Windows for very specific task that I can't do on Arch
Thanks for this. It would have been good to include windows 10 test as well as the people who are thinking of upgrading will get to know the difference
Windows 10 seems faster to me compared to Windows 11.
@@psour33 opposite for me
@@space0015 Did you upgrade from W10 or made a fresh install ? I upgrade here.
@@psour33 On my PC I don't feel a different, but my old Surface feels more fluid with Windows 11
Congrats with being a dad :) Great video!
4:11 is it a bug or is the windows unigine engine running only on 8 threads while it was running on the linux versions with 16 as shown?
Excelent video Bro! and congratulations on your baby!!
Why is everyone hung up on Boot-up Speed times? ?
One note - when you analyze how fast something is happening it would be helpful to see what is load on the system (some sort of process monitor)
Hi, Techy, could you do a video on distros recommended to use with NVIDIA cards?
I'm having a hard time with a rtx 2060 laptop on Fedora 34 with Wayland, whenever I plugin an HDMI cable the whole system freezes, the only way out is the power button :/
Liar boot test, since a unique Windows 11 feature was not used.
The fast boot.
My Windows 11 here starts in 5 seconds after turning on the machine. There's something wrong with this test.
Not to mention, that count for Windows 11, started even before the computer was turned on. Only at 6 seconds, entered the screen where the BIOS of the motherboard is operated and not the OS.
While the Linux distros, they were already on the system loading screen.
This is clearly a manipulation to benefit your personal taste.
He may be disabled fastboot because that feature not available on linux 😅. In my device also linux is very slow in booting
@@thomasrossbabin3671 Windows is better for home users.
@@thomasrossbabin3671 then why do people use bloated os instead of 'perfect' linux os
When Mint gets Wayland it's basically GG for all the distro hoppers out there.
Are no vulkan tests available for graphics benchmarks for linux?
There's a difference in speed between Manjaro and EndevourOS? And if DE's can impact in graphical performance, for example GNOME vs LXDE or something like that.
What file system did Mint and Manjaro use for the bench marking? I wonder if the file system also affects the file read/write and even execution speed.
I don't know about manjaro or fedora but Mint has Nemo as the default file system
Speed test on arch void and windows please bro
I would love to see a deep dive into what plays into these performance differences. For example if you profiled some of these applications to see which calls take longer, what versions of libraries they call into, how much time syscalls contribute, etc.
windows is in the middle for opening big file.:
OS, idling, firefox+libreoffice, open big file
opensusetumbleweed, 1366 MB, +1512 MB, 47 s
ubuntu 23.10, 1720 MB, +1336 MB, 1:11 s
endeavouros, 2220 MB, +1514 MB, 1:21 s
fedora, 2188 MB, +1187 MB, 1:24 s
rocky, 1852 MB, +1724 MB, 1:25 s
mx, 1190 MB, +1884 MB, 1:40 sd
windows 10 2GB SSD, 550 MB, +405 MB, 1:45 s
ghost, 1179 MB, +1139 MB, 1:53 s
linuxmint, 919 MB, +1781 MB, 2:07 s
debian, 2034 MB, +1873 MB, 2:15 s
solus, 629 MB, +1900 MB, 2:29 s
ubuntu 12.04, 1269 MB, +3412 MB, 4:15 s
openbsd, 1329 MB, +1973 MB , 5:08 s
Why not using Vulkan instead of opengl vs directx?
because scared :D
I tried Fedora 34 after it came out and really liked it; however, I couldn't get my system to wake up from sleep. Are others having this problem?
if virtualization is ON on the BIOS a fresh install of windows 11 will activate Core Insolation, that virtualizes the kernel for security, that has a porcentual performance lost, be sure to deactivate it if that's the case (since virtualization is not active in most systems by default)
LOL. ( facepalm)
It doesn't matter if it's virtualized or not....Windows has PROVEN for 3 decades straight to NOT be a secure Operating system.
changing the icy of the cake still means..it's a cake
Now I agree with your point that this virtualization being on in bios CAN hinder performance in gaming and a few things by a small but noticeable enough margin. That's also been proven since winturd 11's launching.
@@motoryzen What is *WINTURD 11?*
Glad I bit the bullet and went down the Mint & Win10 dual boot route a few weeks ago. I'd been on Win8.1 since 2015 but I was increasingly encountering software that had warnings about Win8.1 incompatibility. I'd installed Elementary OS on an old laptop about 5 years ago just to get acclimatised to Linux before letting it anywhere near my desktop, so the learning curve for Mint wasn't too steep. Win 10 stripped of all telemetry, bloatware, and similar nonsense is actually quite beautiful, but to maintain it in that pristine state is too much for me. I want to get things done, not tinker with the OS constantly.
Is it safe to have a laptop with dual boot windows 11 and linuxmint? Will it mess with my windows? Like corrupt it or something ?
@cross_talker Yes, it's completely safe. Just be sure you write to right partitions/drives etc. I'm more worried about linux side, lol. I have read that windows tends to overwrite grub bootloader with its own thus making you unable to boot into mint. Haven't tried dualbooting myself but it should be pretty easy to reinstall grub from any linux usb.
@@the_sweetpotato9254just use a different grub partition and mount windows in Linux so that grub can detect it.
you had a fresh windows 11 install right? at the default there is a feature called VBS security enabled which really make the windows 11 graphical performance worse.
I've done scientific/engineering simulation benchmarks in linux and I've seen fedora peak way higher than arch, so manjaro should be same. All ubuntu derivatives fell behind.
Please do this with windows 11 insider beta they have done insane performance improvements! up to 2x faster startup 4x faster shutdown 2x faster opening apps. This new windows update is insane!
Edit: And remove most telemetry! Would like to see the difference then! For now im keeping notifications on and not switching to a linux distro
Edit V2: I switched to linux anyway 5 months later.
You could test Nobara and different browsers with every system to display which one is the fastest option. The tests felt simple and not really conclusive. Which kernels were the operating systems on? Which version were you using? etc.
Like i wonder why plattform in 4:12 tells, windows 8?! where is windows 11?
Will there be a battery life test? Kind of interested to know how it fare in modern laptop
5:08 - No wonder windows is so bad with all that spyware running in the background xDDD
Good stuff all around, except for testing Kdenlive Nightly on Windows, and choosing Edge (which just barely came out of beta on Linux) for the browser perfomance comparison; both of these favor W11, especially in regards to Edge running on Windows.
Also, no ram usage comparisons?
Ram tests will be in the full article I'm still writing.
Here are the scores:
Windows 11: 316
Linux Mint: 320
Fedora 35: 318
Manjaro: 319
The Geekbench for WIndows was done with the 'Balanced' power plan, I wonder how much better Windows would have been overall with the performance power plan
Arch with KDE can't run smooth live stream in browser + discord while watching game stream + installing some soft. When I run new mid CPU usage task, sound start stuttering while this task loading. Yes. it fast, but looks like not good resource using. I mean, already running task cannot get enough resources for smooth working for a while.
My win10 start discord little longer, or something else, but already working tasks still works as must, until I not start something that use 100% CPU time.
What about doing Windows 10 vs. Windows 11? Be interesting to see if it's worth it.
No
Can you make a tutorial at how to install Davinci resolve and amd opencl on manjaro. I've been try using some package in AUR, and still can't using Davinci resolve
I've been rocking Fedora 35 on Wayland with the Nvidia 495 driver and it's great and buttery smooth! Just make you get the dash to panel extension to make it similar to Windows. People are sleeping on this big time. Actually no, it's literally somewhat of a new thing lmao.
any weird problems with wayland on nvidia? do you think its ready?
@@outofahat9363 as if right now, the implementation on fedora is the only official and native implementation of nvidia on wayland, definitely won't be 100% stable, but from what I heard, it's pretty good already
Why my linux mint boot slow about 33 sec, my laptop specs is amd a8 6410 , 8 gb ram, 128 gb ssd. But if i using windows 8.1 it boot really fast just about 5 sec
Kudos for this! This is one of the few fair "Linux vs Windows" comparisons I've seen on YT. So many of them are so blatantly biased one way or the other, and it's so disingenuous and unhelpful.
So, again... Kudos for keeping it as fair as possible.
I use both systems so I was curious going into this haha
Hey. Can you try disabling core isolation in windows defender and see what results you get?
Now we wonder why Tumbleweed and EndeavourOS were missing in this video...
Can't you run the graphics test on vulkan?
I noticed such a thing. I have Internet 100 mb. on Fedor, the speed test shows 55 and on Manjaro 75. and Fedor in live mode shows 75, but after installing on disk 55. can you do the test
AMD OpenGL performance on windows is bad. That’s why when using directX it was way ahead of Linux. Hopefully AMD address it sometime this century. Vulkan vs DirectX would be good to see
If Linux had 100% compatibility with all games, I would switch to it over windows in a heart beat. I just love the customization, and the privacy.
with the graphics, was Fedora running on Wayland or X?
What about adobe programs? Do you check normal software for works?
Good comparison, but I think you should put the numbers 1 2 3 4 to the corresponding OS to make it easier to see what OS is good in what aspect.
4:10 Windows 11 has been known to be pretty buggy. I wonder if it was Windows 10 instead would Windows had won?
What the hell? My windows 10 installation boots way faster than my linuxmint install on the same machine. How did you get yours to boot in just 13 secs?!
Thanks for making these kinds of videos for us.
KDE for the win. To hell with Windows and GNOME. Great video.
As a huge Microsoft fan, it disappoints me how Microsoft is not doing as well as they did back then, Windows 10 was the best, Windows 11 has serious problems, I'm very impressed with Linux and the way they each performed was really cool to watch
No timestamps? :(
I'll add some. 😎
@@TechHut great
hey techHut can you make a guide on what to do if fedora has very bad video playback? i love fedora but when i used it playing video was just so horrible. i tried the ffmpeg fix the forums said but even that didn't help
not that i am totally taking the L i went on to Manjaro. but i rly wanna give Fedora a chance
I've had great playback in fedora. Are you enabling everything from RPM fusion? Also try MPV for local files.
add few gaming benchmarks like civilization build in turn time benchmark. it feels to me like your blender isnt configured to use hardware acceleration properly
Manjaro KDE Plasma user here! It's awesome! 😁
How do you work on linux with Davinci Resolve? I still have issues with MP4 files and need to convert them. That sucks as hell. Any recommendations?
I don't like file transfer animation in gnome. It doesn't provide any information about file transferring or not.
Could we get the Linux kernel versions next time? Newer versions have had some optimizations for AMD CPUs.
Nice video, man!
We have Edge on all this OS, how about test with Firefox?
Huh?
Remember that GIMP is based on the GTK framework and is not easily compatible to Windows and has to go through a built-in compatibility layer of some kind.
Comparing games by average FPS is really stupid way to do it. 1% lows is the important value since that is the most distracting aspect, the FPS dips and frame times is what we are sensitive to.
Yes..but that DOESN'T show how often within any individual game that it DIPS to that lowest/worst case fps......THAT is the most critical deal followed by what fps the game MOST OFTEN runs at.
Why do you use Edge instead of Chrome in web benchmark? I believe Edge is unoptimized in Linux. I tested Chrome base in Linux that Edge has the lowest score compare to Chrome base such as Chromium or Chrome.
the only reason windows was faster in your kdenlive test is probably because you used flatpak. A native package will definitely squeeze out more performance.
I'm going to be taking a deep dive. Looking to this in an upcoming video. :)
@@TechHut YESSSSSS
now we will see the TRUE power of Linux so excited
Why did you change the thumbnail?
vulkan instead of opengl?
Shouldn't you test linux vs linux kernel version then fastest kernel verses windows?
I'm done. If you don't understand why Windows would beat out flatpaks, you haven't used Linux enough. KDENLive is available (native) in pretty much all distros. It's just weird that you would use what essentially a Java KVM to run against Windows native.
Needed to open the same exact project requiring the same exact version on 4 different platforms. Flatpaks are the direction we're going in general so I figured it was a fair test. Also watch out for the video I make on this topic. Rendering doesn't give you that much of a difference compared to installing directly from the repo.
@@TechHut I will be honest and say that USING programs in flatpak form is not something I have a lot of experiene with...so, I am looking forward to the rendering. But, since flatpaks are the basic equilavent of running programs in a virtual environment, I stand by my objection using them to determine load times... cuz, quite obviously, the virtual environment has to load first.
If it works better (or even comparable) inside that environment, fine! I will accept that.
what version of Mint you were using?
Do FPS against Vulkan games and DirectX games
Very uncool that you didn't run Vulkan on Linux like you did with DirectX on Windows. I'd expect better from a channel mostly about Linux.
These benchmarks are to taken with a grain of salt,, Different kernels, latency settings in the linux distro's for a start, The file copy test is out, as linux writes to ram first then disc.
Thanks for the comparisons with my favorite distros!
A lot of these were really hardware benchmarks, like grinding on the CPU to encode a video, the OS doesn't affect that much. I'd like to see some benchmarks that cover the human usability stuff, like in Windows if you open your downloads folder with like ~100 files in it, it takes multiple seconds for the directory contents to appear, making the OS feel very slow in day to day use. Also the amount of time it takes for the right click menu to appear. The Windows start menu is also horribly slow, often showing blank UI for a few frames before text labels appear. Launch times for various programs would be more useful than boot time (since I rarely reboot), especially when you have a few programs running already.
Hi, I have an old dual-core pc higher than 2ghz and 2 giga ram and a minimum gpu I'm interested in the best programs for photo and video editing which distro do you recommend for more versatility? Thank you.
Can you beat 11 sec. from grub to desktop with kubuntu 20.04 LTS, Intel i5-i3330 and samsung SSD on a ex-gamer PC?