The main thing is not the object of measurement, but the measurement technique. You can always choose the measurement method so that you get the results you want. Living in Russia, I know this very well.
@@azulfenix7 Windows is used by like 95% of people. It’s definitely good comparing against it, especially for people who are considering trying out Linux and maybe switching to it
An important thing to note about the game benchmark is the fact almost all games use DirectX not OpenGL. I find that DirectX to Vulkan translation costs about 10%-15% of your overall FPS in DX11/12 games, so unless the game natively supports Vulkan or OpenGL it will perform worse in Linux. World of Warcraft actually performs the same in DX11 mode as it does in DX12 mode which is a shame as DX12 doubles the frame-rate for me on my laptop in Windows. Otherwise this is consistent which almost every game I tried.
but aren't like... a huge % of games written using engines like unreal, cryengine or unity? all of which have vulkan ports (i assume... they all work with consoles!)
Just as additional clarification, a backdoor is specifically code that has been included in a piece of software, either by the developer or by someone who gained access to the source code, to allow someone who knows about the backdoor to circumvent the normal, readily apparent security measures.
@@CFWhitman For additional clarification next time fill up the bugs report. #Bug Report ==================== Bug ID: 'cQWwb8VRrhU' Bug Name: 'Backdoor definition issue.' Summary: During video recording which later published to "RUclips" platform (tags used: "#TechHut #Windows #Linux", publisher: "TechHut" date: "3 Jan 2021" link: "ruclips.net/video/cQWwb8VRrh/видео.html> Submit Date: 17 Jan 2021 URL: 1* Backport: en.wikipedia.org/wiki/Backport 2* Backdoor: en.wikipedia.org/wiki/Backdoor_(computing) Expected result --------------- 1* Backport Actual result ------------- 2* Backdoor Screenshot ---------- None Notes ----- The bug first reported by: Hamlet Elsinore. ==================== #End report
Appreciate the review. You did lightly mention a kernel update. I would focus a little more on the kernel versions on each or state them in the beginning because, as you know, the kernel and modules have a lot to do with the performance of the system and optimization of the hardware.
@@doge3911 from the endpoint of your sentence you demonstrate that you have no clue what you are talking about. Kali Linux users have fallen off their chairs laughing. Windows better for hacking? In a dreamworld, maybe. As for the gaming now. What are you? 4 years old? A computer is primarily a work machine for the power user. If you want a gaming machine, go buy a console. If your argument is that Windows is a better OS, just because it games, I, now, started laughing. Try running it on a server or in scientific projects. Even Microsoft knows it is trash, that's why Azure runs Linux. As the video clearly showed you, Linux distros are far more efficient than your locked down personal information transmitter. Oh, I meant OS...
@@doge3911 For some reason I get more fps on linux with a broken nvidia driver, than on windows 10 fully working... Strange. All linux native games run 10% better btw
I agree. I guess many of libraries required by the application were already loaded on GNU/Linux because they may be used by the rest of the system. I think this is why the application loaded a lot faster. I think he should also compare performance of GIMP when it is loading from AppImage or use statically linked applications.
@@nymusicman I just tried to run GIMP 2.10.12 myself on the same version of MS Windows like he used. It loaded in 7 seconds ("Loading brushes..." step took the most of the time). Then I tried to do a reboot and try to run it again just after MS Windows loaded. Then it loaded in 14 seconds (different step took most of time, I don't remember which one but not fonts). I guess that just after reboot it takes more time because MS Windows still loading other stuff in the background. Then I tried to run GIMP on LUbuntu with LXDE (on the same PC), I didn't check which version. GIMP loaded in slightly more than 2 seconds. My PC is almost 10 year old Dell Vostro 3350 with Intel Core i5-2430M, 16 GB DDR3 RAM, SSD on SATA 3. So it seems very strange to me that his PC took over 30 seconds to load GIMP.
I would like to see this test performed on a machined that's been in use for a year or so, running the typical everyday apps. Windows is fast on clean install but after a few months drag to a halt. I use a Debian based distro which doesn't appear to do the same and all my apps are still rather snappy.
It's been over a year, so I think it's mandatory you re-run these with latest versions (Windows 11/Debian 11) and maybe also throw OpenSuse Tumbleweed into the mix to represent Slackware.
Nice work. I've settled on Debian, though I think for specialized purposes, Arch/Gentoo/LFS with purpose specified compiler flags will always be impossible to beat. My only use for M$ is to play games. Thank you again, for the review.
Agree. Support, documentation, compatibility.... Many things to consider on top of performance. If you need 3 days to figure out something in your Distro, the total performance is not that great right?
@@maxc7198 M$ makes money by letting people pretend to be smart, but not really know what is going on, let alone do real things about it. That's the way of closed source propriety-ware. Free and open ware on the other hand, is just that, free and open. Want more? Do more. Want less, do less, or take it as it is. What I like best about GNU/Linux is that I'm working with good people, not bad people. I choose not to work with bad people, because I like myself that much. :-)
@@davidweeks1997 Good point. Can be traced back to Stallman. But mainly ppl does not want to know. If you become aware, you know what you support (ex. Amazon). Same with software, privacy and security.
Moreover, the linux kernel shipped in Debian is very generic, and has no specific laptop/desktop optimized flavour. I use Debian, and performances increased significantly after I configured and built a kernel with more recent source code and a leaner configuration.
It probably played a big role in the PHP Benchmark. The advantage of Arch is the newer software (this is not always an advantage). I don't think that it is more light weight, Debian and Arch are most likely the same when it comes to system recourses.
its because Debian doesnt update so often, you will see old software version, for example, if Windows, Ubuntu o any other has Firefox 54, you will see in Debian Firefox 52 kinda. They wont update till they are complete sure that is safe for your system.
3 года назад+1
@Richard Vaughn Maybe testing instead of stable would have been a good choice.
Nice comparison, much appreciated! Anyway, I will continue to use Debian SID as my "daily driver" :) I don't recommend it on a server but it is very stable for a daily usage on a laptop or on a desktop and it is a "rolling" release like Arch. Peace!
I use OpenSCAD, win10 122 seconds, Arch 60 seconds, Debian 57 seconds on a particular design I tested. It really does seem that the specific tasks really matter.
I just gave Manjaro a try on an old laptop I use every once and a while. I didn't realize how snappy and fast arch was on it compared to the other debian OSs I have been using on it.
I have a netbook that Ubuntu (Unity DE) absolutely crawled on, but Lubuntu (selectable at boot screen) worked just fine. Other distros of the era didn't enable the netbooks wifi card OOTB, so I was forced to stay with it. Aside from the annoying fact that some settings were not accessible via the Lubuntu (or vanilla LXDE Canonical also offered), forcing me to occasionally boot into the molasses-like Unity DE.
@@vinyl.croatia Considering M.2 are getting cheaper SSD's are not really "high end" now, I got it for $99. I know Linux is great but even when I had a trash HDD my windows booted in only 15 seconds or even less.
So I did a similar test before between windows 10 and Ubuntu, for GPU rendering. Dual boot so both on the exact same system off the same drive. Windows perform 24% better in its ability to use the GPU for rendering 3d video. This is something I specifically have to do so it matters to me, especially when a render might take several hours- 1/4 saved time means a lot. Of course this is a very specific usage and doesn't really represent normal use, but it's not what I expected... especially considering that both operating systems were using the official nvidia drivers.
gimp and libre are so slow on windows has to load more libs or something that linux already has my guess. gimp feels emulated on windows these days. definitely one of my favorite apps to work in. i love arch with gnome so refreshing when you open something you can really tell it throws your full power at the task you give it while windows has to queue up the spying of that app and most likely its massive code length just to open anything before opening it. once youre inside of a aaa game on windows its pretty alright i guess and has full backing of nvidia, steam and any other corp making the title specific to the os, kind of unfortunate that linux doesnt have windows level of a game library but its still good enough that id like to make the full switch soon, right now i just have arch on a usb 2tb id like to put it on a m.2 ssd when i get one.
Yea Linux, especially Gnome, which is based on GTK (Gimp ToolKit), uses shared libraries that are used by multiple programs. When you run Gimp, the GTK libraries are already loaded, because the whole desktop environment uses those libraries, and Gimp uses those instead of loading new ones. That cuts the loading time a lot. On Windows, every time you run Gimp, the GTK libraries have to be loaded.
Interesting. I wonder what causes differences. I assume that graphics drivers are slightly different, memory manager is different, and running background tasks are different. For me most important would be something like application tests: time to run gimp, blender, to open some big file, to find specific file inside directory structure, to perform maintenance/optimization of large database etc.
Probably would have been a more fair comparison between distros if you ran the same desktop environment between the two since their overhead will probably have more of an impact on performance than anything else. Though you do also have windows as a contender, so hey, it's already apples to oranges. If anything it would be nice to know the desktop environments you used for each OS, I recognize Gnome on Debian, but I can't discern what DE you're using for the endeavouros machine.
@@escapetherace1943 Bro wasn't even using real Arch. True Arch would have performed better on most except probably the render test, because he used EndeavorOS, which is optimized for gaming.
@@shringe9769 arch literally sucks and is cringe and populated by people who think they are powerusers when their programming socks are on too tight no thanks
Endeavour os is an absolute banger of an OS and a great implementation of arch. Super stable, literally the same as arch in the sense of a vanilla arch installation, super clean,super snappy
You could try Debian vs. Devuan with similar desktops to see how much time SystemD adds to the boot process, by my experience it's a surprisingly long time. And also we would see how much overhead is added.
@@lawrencedoliveiro9104 ? I didn't say I did. I said I had a given experience, you answered you had good reasons to think that it ought to be the opposite, and finally i answered that it was all the more reason to try to put it to the test.
Nice video. I watched the older ones too. It would good if in comparison videos is mentioned and the power management (battery life) during the tests especially for these on laptops. I thing it would be helpful do userts to choose between a second or two of performance and a minute or two to use their machine. ✌🏻
Boot speed was not fair, I think the desktop environments should be same on both the distros, i.e. the lightest Desktop available and should be compared to the lightest version of windows wiht no unecessary apps on startup. Good Effort overall, keep up the good work.
Hello TechHut, thank you very much for the great content, I have a question about the camera you are using for your RUclips channel? If you can share the details I will really appreciate it! Keep the great work!
Arch also has different kernel versions, one of them is Zen Kernel which is an optimized version supposed to be "the best" in terms of performance. You should try this on next.
The best in performance will always be the one tailored & built for your specific hardware and cpu, I use zen because of the features, I can't bother to test performance
Hey TechHut, how about a challenge: Using Raspberry 0 w as your main PC for a day or three? I'm sugesting this because in 2012, Chris from Explaining Computers used Rpi2 as his PC for a week. Now in 2021, covid made more jobs change to the online model, and here in Brazil the prices of computers got very high. As i'm using my Rpi4 for main computer for a whole year, ive had a really good experience on manjaro with sway etc. Many wonder what can be used or not in Rpi0w, so this test could be great. Some details to make the sugestion working: -SYSTEM: raspberry OS. Maybe you get a better experience with arch, dietPi or even Gentoo idk. -DESKTOP ENVIROMENT: change the LXDE to sway or i3 (novaspirit has a great tutorial on this). Maybe 540p resolution would be better -PROGRAMS: abiword instead of libreoffice, gimp?, firefox with lightweight extensions or some terminal browser (i can sugest more tips on this topic) -HARDWARE RECOMENDATIONS: SSD or HD with external powering instead of the microSD memory, a good microSD or even leepspvideo's PiSafe's eMMC to boot, keyboard with dongle, cable management, and of course a great usb hub or the GPIO shild with more USBs -CHALLENGES: try typing some text, light image and audio editing, youtube playback with h264ify or pasting the link in vlc, gaming (even light PS1 games works), benchmarks, light versions of sites like gmail and facebook, video or audio conferences, multitasking in general with htop, wine, DRM, ssh, cmatrix of course, etc
I wonder if different linux desktops would have different results in 3D benchmark? I once did the Half-Life 2 Lost Coast benchmarking using ArcoLinux with different desktops and had interesting results.
I'd be curious to see the key libraries/package differences aside from the kernel that might be contributing to the differences in the debian and arch system. also, if there were any driver/cpu microcode version differences, and any differences in the default cpu power settings for the distro or if change the amd processor profile is different in different distros.
2:14 - Windows launched it so slow because of the Windows Defender I'm 99% sure. Edit: I also found that Linux have loaded most of the libraries (they are needed for DE) while Windows does not (it uses Win32 API and not GTK+)
The biggest Speed gain is when actually using the OSs .. Any Linux Distro will smoke Windows .. The system lag with Windows is massive, application crashes, update hell and virus issues when using Windows. Open Source FTW !
@@Jameshazlett As I said in a another comment "I've done other benchmarking videos with GNOME and XFCE distros. The difference is so minor in tests like this unless you're running a really low end system. The DE might add half a second to boot and opening the file manager may take a tenth of a second longer to open. I went with that what the distros default DE was. "
Curious to see if Debian with updated Mesa would've scored more on the GPU test. That's a fun test you've done, thank you. Do you think Ubuntu would've been more appropriate? Asking because Debian's philosophy is focusing on rather outdated, albeit well-tested, packages and none of the other two OSs in the test are *that* stale in terms of package version.
At home I have an old HP stream 13 netbook of 2015. It still works nicely with MX linux on emmc internal drive and also manjaro on a micro sd card. Of course I use light desktop environments such as lxde and xfce... With installed windows 7 I remember impossibility to watch full hd videos. Indeed on Linux with mpvplayer I can watch whatever. Great linux!! How do you score my English I posted from Italy?
Pick your OS (poison) for the need. Those that like Windows will never switch to Linux; either due to feeling it is too complicated or due to software compatibility. As for Linux, I dual boot (separate drives) as I’m a Windows support / system admin and that is the corporate environment I need to keep up with, but I really like my Arch install. To me, Arch just flows so much better than Debian based systems. Here again, there’s Debian fan boys that will scathe my comment; and Arch users that will praise me for running vanilla Arch rather than Endeavor or Manjaro. Which is another reason why some won’t switch to Linux. The animosity and outright hostility towards Windows is one thing; distro to distro is just sad for the community at large.
ya, compatibility... speed is fine, but if i cant do what i want to whats the point.... first software i install are: key manager, XmouseButtonControl, hot virtual keyboard, directory opus, lightscreen, 7 stickyNotes, MT5/MLQ5 trading, potPlayer, notepad++, NetLimiter 4, Tixati, sordum topmost, need the magnifier, +++... what am i gonna do with any Linux???... lol
Well, it's very interesting, but I'm not sure how would it turn out on completely different hardware. I mean - the speed of various tasks may strongly depend on the driver software. Some drivers might work faster on one OS and slower on another. So - on those laptops Windows is slower. But I'm not sure, maybe there are hardware configurations that would give completely different results. Anyway, thanks for the test, it was very interesting to watch. BTW, I remember the times the Linux struggled with 3D graphics. The times of very poor GPU drivers. I see it changed a lot.
@@TechHut Yes, but packages on debian are really old and sometimes makes performance bad, ubuntu and pop are far superior for people who needs performance..especially for gaming.
wow a comparison we actually care about. I do wish there were more real world tests. opening browsers and other software that people use often. Now you need to do all these test in a Garuda system with the Zen kernel.
I've done other benchmarking videos with GNOME and XFCE distros. The difference it's so minor in tests like this unless you're running a really low end system. The DE might add half a second to boot and opening the file manager may take a tenth of a second longer to open. I went with that what the distros default DE was. 🤓
@@TechHut does debian have a default de? I am using it and it just asks me which de to use during installation. How do you know that gnome is default for debian
Default is the wrong word. It's the first option and it's what they commonly use in demonstrations. The main point of what I said was the lack difference on systems with okay specs.
@@cevizagaci31 In Debian, if you select "Debian desktop environment" during installation, and you don't actually select which environment below, then you will end up with GNOME anyway, so, technically, it's the default, though a very weak default since it asks you what you want. This also brings up another point, though. I believe that Debian with GNOME uses Wayland by default, so that could be another difference in the general environment being used.
Windows runs smoothly in a fresh install, a year later you feel the need to buy a new computer, debian/arch runs smoothly in a fresh install and a year later runs better lol
@@laurentiucalonfir Phoronix does many more tests and more or less Debian is in neck to neck with Arch. With wild variation of Arch as it has regular changes.
Booting into OS is NOT a benchmark. It is NOT a process. It's a random collection of custom services YOU DECIDE TO START. Similar with "starting application". It's NOT a benchmark. Do you use caches? Prefetch? etc. etc. etc. Slower might be better, and might be not. Depends.
"YOU DECIDE TO START" this was the base installation of all of these operating systems with nothing changed other than the login requirements for all of them.
@@user-dc9zo7ek5j it's a desktop. From the time I power on at the back of the psu, turn the machine on and sit down I am on the desktop. Also fast boot in the bios isn't active.
@@SleepingSavage Well then, 7s from totally turned off is really fast. My linux boots for 16 seconds, with all the bios logos and the grub delay. And I did not mean fast boot in bios, I meant fast startup, it's a special windows feature after win 8, that improves the startup by actually not shutting down fully.
@@user-dc9zo7ek5j yep I'm still surprised by how quick it is. Fast start up is not active. Hibernation states (of any kind) are useful on laptops but personally I never use them on anything. Just a personal preference.
@@kafeshop Oh im sorry, i didnt realize that you are one of the dudes that seriously believes that debian and ubuntu are the same... Or even better that all linux distros are the same. Lol. Educate yourself and make your own research. ;) Its like you are saying that a VW Golf and a Seat Ibiza are the same car only cause the engine is common on both cars. Nope. :)
@@kafeshop Well, when you are a professional and not a hobbyist and you need something to work perfectly out of the box, you give some extra attention to the details of each distro. No no man, its not about fashion, seriously its obvious that you are using linux for fun and not for productivity.
@@orkhepaj because it compares how both base systems behave and which system is better for which activity and to mesure pure performance, as arch and debian are very different. The derived distributions modify which services and programs run in the background and desktop environments have an important impact on performance, thus makes the comparison very unfair and very difficult, and what adds to that is the fact that he used vanilla debian and not vanilla arch linux but a lightweight distribution based on it, it is like comparing a plane with a car. I think he should compare either the base systems, or a lightweight distributions of BOTH systems
@@nournemroud plane with a car?:O nah these tests should be done on used and tailored systems , like what a normal person would have after a month or two
Windows for me Never had much good experience with Linux in general.. Ubuntu was the best but still had a few issues like screen tearing even tho latest drivers were installed and vpn not working Debian didn’t boot at all every time after installing 4 times in a row OpenSuse was good.. but many key features were not working like simple video playback, suspend issues and again screen tearing And for me it’s not worth to invest hours to fix this Linux as Main OS is cool if you have the time to work everything out .. but for productivity it’s just not near reliable as Windows is I know those experiences are subjective, just my opinion Also one example: I created a windows installer on USB one year ago and this works no matter how often I use it Making a USB installer for a Linux distro has some luck to it .. sometimes it works, sometimes it reports errors during installation
I don't know why but whenever I boot linux on a separate drive or an usb stick, something happens to windows and cpu usage goes down. For example, before I used linux I had 15 percent cpu usage in windows. After using linux, it went down to 7 percent. Now it is at 3 percent. That concerns me. As for your post, linux works fine so does windows. I have no problem using both
The main thing is not the object of measurement, but the measurement technique.
You can always choose the measurement method so that you get the results you want.
Living in Russia, I know this very well.
In America you control measurement method - in Russia Measurement method control YOU.
> In America you control measurement method
@@EddieVBlueIsland, how presumptuous!
@@Михаил-н4е4ъ you really think this is happening only in russia !
@@ahmedegymed5853 no, i don't think so. just in russia nobody has illusions about this.
@@Михаил-н4е4ъ 👍
Booting strats*
Arch: "I'm gonna give you guys a little advantage but after that i'll not hold myself back "
arch => one way ticket :)
The truth is, I don't know what windows does here, to begin with, you already lost simply because Linux is not only great, but it's also free
all good till you get inside FreeBSD, then any Linux distro is done.
@@carzoparazzo9698 what are some good Unix distros ?
@@azulfenix7 Windows is used by like 95% of people. It’s definitely good comparing against it, especially for people who are considering trying out Linux and maybe switching to it
An important thing to note about the game benchmark is the fact almost all games use DirectX not OpenGL. I find that DirectX to Vulkan translation costs about 10%-15% of your overall FPS in DX11/12 games, so unless the game natively supports Vulkan or OpenGL it will perform worse in Linux.
World of Warcraft actually performs the same in DX11 mode as it does in DX12 mode which is a shame as DX12 doubles the frame-rate for me on my laptop in Windows. Otherwise this is consistent which almost every game I tried.
but aren't like... a huge % of games written using engines like unreal, cryengine or unity? all of which have vulkan ports (i assume... they all work with consoles!)
@@cho4d maybe in the future but still most games don't support vulkan by default
09:25 not backdoor but backports (backdoor is the term that describes a security hole, exploit).
Thank you, my apologies.
Just as additional clarification, a backdoor is specifically code that has been included in a piece of software, either by the developer or by someone who gained access to the source code, to allow someone who knows about the backdoor to circumvent the normal, readily apparent security measures.
@@CFWhitman For additional clarification next time fill up the bugs report.
#Bug Report
====================
Bug ID: 'cQWwb8VRrhU'
Bug Name: 'Backdoor definition issue.'
Summary: During video recording which later published to "RUclips" platform (tags used: "#TechHut #Windows #Linux", publisher: "TechHut" date: "3 Jan 2021" link: "ruclips.net/video/cQWwb8VRrh/видео.html>
Submit Date: 17 Jan 2021
URL: 1* Backport: en.wikipedia.org/wiki/Backport
2* Backdoor: en.wikipedia.org/wiki/Backdoor_(computing)
Expected result
---------------
1* Backport
Actual result
-------------
2* Backdoor
Screenshot
----------
None
Notes
-----
The bug first reported by:
Hamlet Elsinore.
====================
#End report
Wanted to like your comment about back doors but it already has 69 likes. Blessings upon your bootyhole
@@thebedroomintellectual2460 what?
Thanks for the video! I have one suggestion: keep all three charts with the same maximum and minimun axis values to better give a notion of proportion
Fr
Just some constructive criticism: the graphs would be much easier to interpret visually if they had a matching scale
Appreciate the review. You did lightly mention a kernel update. I would focus a little more on the kernel versions on each or state them in the beginning because, as you know, the kernel and modules have a lot to do with the performance of the system and optimization of the hardware.
Yeah should have shown more of a spotlight on that from get go. Next time for sure. Full details of OS versions are in the description.
@@TechHut Did you update arch as well? The latest kernel is 5.10.x and you had 5.8.10 (you can see it in at 3:33)
@@doge3911 It's not April Fool's anymore mate...
@@doge3911 from the endpoint of your sentence you demonstrate that you have no clue what you are talking about. Kali Linux users have fallen off their chairs laughing. Windows better for hacking? In a dreamworld, maybe.
As for the gaming now. What are you? 4 years old? A computer is primarily a work machine for the power user. If you want a gaming machine, go buy a console. If your argument is that Windows is a better OS, just because it games, I, now, started laughing. Try running it on a server or in scientific projects. Even Microsoft knows it is trash, that's why Azure runs Linux. As the video clearly showed you, Linux distros are far more efficient than your locked down personal information transmitter. Oh, I meant OS...
@@doge3911 For some reason I get more fps on linux with a broken nvidia driver, than on windows 10 fully working... Strange. All linux native games run 10% better btw
Arch with KDE Plasma for life.
What a soyboy
@@8w73 how come? I'm a dev yeah, but not a soydev ;)
Hell yeah
Arch + XFCE on Mac Book Pro 2015. Big Sur ! 😁 What's that ?
Arch + Xmonad lol
Doesn't GIMP on Windows require compatibility libraries? I think this is not a good app to use for comparison.
I agree. I guess many of libraries required by the application were already loaded on GNU/Linux because they may be used by the rest of the system. I think this is why the application loaded a lot faster. I think he should also compare performance of GIMP when it is loading from AppImage or use statically linked applications.
Yes its true, those libraries are included on the installer..
In general, GIMP takes forever to load fonts on Windows. The UI tells you where GIMP is getting stuck and on Windows it's always fonts.
should compare photoshop on windows and linux....
@@nymusicman I just tried to run GIMP 2.10.12 myself on the same version of MS Windows like he used.
It loaded in 7 seconds ("Loading brushes..." step took the most of the time). Then I tried to do a reboot and try to run it again just after MS Windows loaded. Then it loaded in 14 seconds (different step took most of time, I don't remember which one but not fonts). I guess that just after reboot it takes more time because MS Windows still loading other stuff in the background. Then I tried to run GIMP on LUbuntu with LXDE (on the same PC), I didn't check which version. GIMP loaded in slightly more than 2 seconds.
My PC is almost 10 year old Dell Vostro 3350 with Intel Core i5-2430M, 16 GB DDR3 RAM, SSD on SATA 3. So it seems very strange to me that his PC took over 30 seconds to load GIMP.
Aaaand you did it again. Many thanks, just what I was looking for (again).
I would like to see this test performed on a machined that's been in use for a year or so, running the typical everyday apps. Windows is fast on clean install but after a few months drag to a halt. I use a Debian based distro which doesn't appear to do the same and all my apps are still rather snappy.
If you have any dev tools, windows will poop after a few months. I literally keep a USB ISO to repair or reinstall windows. Its sad
Yes because of all the temp files left over bloatware if you have any etc...
7:20 Usual reddit help, people immediately instead of being helpful going "wtf why would you do that!!!???" like a crazy person
yes reddit is so toxic you cant get a clear answer
It happens on Quora too. You get called a simpleton and a long rhetorical explanation on anything but the damn question itself.
I've been using Endeveaour OS for months with Kde and I'm fine with it, even if I miss some Windows software
KDE is both the most customizable DE and the most noob friendly. It's awesome.
windows is the same
It's been over a year, so I think it's mandatory you re-run these with latest versions (Windows 11/Debian 11) and maybe also throw OpenSuse Tumbleweed into the mix to represent Slackware.
Or just Slackware to represent Slackware. It got an update this year!
You need SSD bro. Linux distro will run faster compare to any Windows system. Additionally, the testing will give you better results.
Wow! Very thorough comparison you did there.
Must have taken forever to complete.
Thank you, it's very informative. Instant sub.
Please use consistent scales in the graphs when comparing
Nice work. I've settled on Debian, though I think for specialized purposes, Arch/Gentoo/LFS with purpose specified compiler flags will always be impossible to beat.
My only use for M$ is to play games.
Thank you again, for the review.
Agree. Support, documentation, compatibility.... Many things to consider on top of performance. If you need 3 days to figure out something in your Distro, the total performance is not that great right?
@@maxc7198 M$ makes money by letting people pretend to be smart, but not really know what is going on, let alone do real things about it. That's the way of closed source propriety-ware. Free and open ware on the other hand, is just that, free and open. Want more? Do more. Want less, do less, or take it as it is.
What I like best about GNU/Linux is that I'm working with good people, not bad people. I choose not to work with bad people, because I like myself that much. :-)
@@davidweeks1997 Good point. Can be traced back to Stallman. But mainly ppl does not want to know. If you become aware, you know what you support (ex. Amazon). Same with software, privacy and security.
@@davidweeks1997 thats probably the most mindless self absorbed comment i've seen all day
Debian has older versions of the same applications that Arch has. Could this play a role in the outcome of the tests?
Moreover, the linux kernel shipped in Debian is very generic, and has no specific laptop/desktop optimized flavour. I use Debian, and performances increased significantly after I configured and built a kernel with more recent source code and a leaner configuration.
It probably played a big role in the PHP Benchmark.
The advantage of Arch is the newer software (this is not always an advantage). I don't think that it is more light weight, Debian and Arch are most likely the same when it comes to system recourses.
I also not convinced that he used the same exact laptop for every test unless he's got heck alot of money
its because Debian doesnt update so often, you will see old software version, for example, if Windows, Ubuntu o any other has Firefox 54, you will see in Debian Firefox 52 kinda.
They wont update till they are complete sure that is safe for your system.
@Richard Vaughn Maybe testing instead of stable would have been a good choice.
Nice comparison, much appreciated! Anyway, I will continue to use Debian SID as my "daily driver" :)
I don't recommend it on a server but it is very stable for a daily usage on a laptop or on a desktop and it is a "rolling" release like Arch.
Peace!
all of them won and lost some tests. but when Arch loses it's never a big difference. it was the most consistent
Yeah
It's not good enough to leave Windows's user friendly interface tho.
@@serhafiye7046 cringe
@@michaelwu7419 how
@@serhafiye7046 really? install gnome or cinnamon on arch and the UI is more friendly than that of windows'
I use OpenSCAD, win10 122 seconds, Arch 60 seconds, Debian 57 seconds on a particular design I tested.
It really does seem that the specific tasks really matter.
I just gave Manjaro a try on an old laptop I use every once and a while. I didn't realize how snappy and fast arch was on it compared to the other debian OSs I have been using on it.
and buggy.. manjaro xD
Congratulations for the video!
Maybe the Arch results would be even better using the original Arch, not an Arch based distriburion.
Windows is fast, but only on really high end hardware, while linux is fast on any machine.
@@pikachulovesketchup666 Windows is not bad, its a good operating system, but linux is really superior in many aspects..
I have a netbook that Ubuntu (Unity DE) absolutely crawled on, but Lubuntu (selectable at boot screen) worked just fine. Other distros of the era didn't enable the netbooks wifi card OOTB, so I was forced to stay with it. Aside from the annoying fact that some settings were not accessible via the Lubuntu (or vanilla LXDE Canonical also offered), forcing me to occasionally boot into the molasses-like Unity DE.
I don't have a high end PC, just a good SSD and windows boots faster than my monitor turning on...
@@arieljgs Well, that SSD must be high end then...
@@vinyl.croatia Considering M.2 are getting cheaper SSD's are not really "high end" now, I got it for $99. I know Linux is great but even when I had a trash HDD my windows booted in only 15 seconds or even less.
So I did a similar test before between windows 10 and Ubuntu, for GPU rendering. Dual boot so both on the exact same system off the same drive. Windows perform 24% better in its ability to use the GPU for rendering 3d video. This is something I specifically have to do so it matters to me, especially when a render might take several hours- 1/4 saved time means a lot.
Of course this is a very specific usage and doesn't really represent normal use, but it's not what I expected... especially considering that both operating systems were using the official nvidia drivers.
gimp and libre are so slow on windows has to load more libs or something that linux already has my guess. gimp feels emulated on windows these days. definitely one of my favorite apps to work in. i love arch with gnome so refreshing when you open something you can really tell it throws your full power at the task you give it while windows has to queue up the spying of that app and most likely its massive code length just to open anything before opening it. once youre inside of a aaa game on windows its pretty alright i guess and has full backing of nvidia, steam and any other corp making the title specific to the os, kind of unfortunate that linux doesnt have windows level of a game library but its still good enough that id like to make the full switch soon, right now i just have arch on a usb 2tb id like to put it on a m.2 ssd when i get one.
Yea Linux, especially Gnome, which is based on GTK (Gimp ToolKit), uses shared libraries that are used by multiple programs. When you run Gimp, the GTK libraries are already loaded, because the whole desktop environment uses those libraries, and Gimp uses those instead of loading new ones. That cuts the loading time a lot. On Windows, every time you run Gimp, the GTK libraries have to be loaded.
@@ClifffSVK Even if you have multiple Windows apps that run on GTK, those libs are independent of each other and has to be loaded separately.
I like these comparisons. I'd be interested to see chrome OS in comparison to these as well.
Thanks for your considerable efforts on this one. 👍
Interesting. I wonder what causes differences. I assume that graphics drivers are slightly different, memory manager is different, and running background tasks are different. For me most important would be something like application tests: time to run gimp, blender, to open some big file, to find specific file inside directory structure, to perform maintenance/optimization of large database etc.
Probably would have been a more fair comparison between distros if you ran the same desktop environment between the two since their overhead will probably have more of an impact on performance than anything else. Though you do also have windows as a contender, so hey, it's already apples to oranges. If anything it would be nice to know the desktop environments you used for each OS, I recognize Gnome on Debian, but I can't discern what DE you're using for the endeavouros machine.
How can you ask that of an arch fanboy though? He clearly is trying to make arch look best, as usual
@@escapetherace1943 Bro wasn't even using real Arch. True Arch would have performed better on most except probably the render test, because he used EndeavorOS, which is optimized for gaming.
@@shringe9769 arch literally sucks and is cringe and populated by people who think they are powerusers when their programming socks are on too tight
no thanks
It'll be nice to see the results with debian testing instead of stable :)
Arch with Alacritty is the fastest terminal setup I’ve ever used. It feels instantaneous when starting a terminal window.
mine debian+st
Very interesting comparison.
Thanks for posting.
Endeavour os is an absolute banger of an OS and a great implementation of arch. Super stable, literally the same as arch in the sense of a vanilla arch installation, super clean,super snappy
I just checked the website and I can't even find where the download gnome version ?
@@buleulek5153 It hasoffline and online installer, you can choose DE at time of installation (with online installation)
About the Unigine benchmark, AMD drivers on windows suck at OpenGL so that's one of the reasons it's has a worse score.
+1
You could try Debian vs. Devuan with similar desktops to see how much time SystemD adds to the boot process, by my experience it's a surprisingly long time. And also we would see how much overhead is added.
systemd runs boot tasks in parallel where possible, which makes them faster.
@@lawrencedoliveiro9104 In my experience every time i try a non-systemD distribution the boot is radically faster.
@@themroc8231 Why do you think others tend to have quite a different experience?
@@lawrencedoliveiro9104 But this little controversy serves to prove my point: it would be interesting to try.
@@lawrencedoliveiro9104
?
I didn't say I did. I said I had a given experience, you answered you had good reasons to think that it ought to be the opposite, and finally i answered that it was all the more reason to try to put it to the test.
Nice video. I watched the older ones too. It would good if in comparison videos is mentioned and the power management (battery life) during the tests especially for these on laptops. I thing it would be helpful do userts to choose between a second or two of performance and a minute or two to use their machine. ✌🏻
Do openSUSE next. I was testing it on my laptop versus others. Booted faster, at least for me, than Debian and Arch based.
@techhut, you did a miscalculation at 10:17 .... Debian is ahead by 9 seconds compared to Arch OR Debian is ahead by 6 seconds compared to Windows!
Thank you so much !
Please make competition with Linux solus 🙏
Thanks for watching! I have a video covering Solus vs Ubuntu.🐧
techhut.tv/post/2020/11/solus-vs-ubuntu-budgie/
Great idea...This comparison is rarely compared the base OS distros && Windows. Cool!
Wow interesting! Many thanks! Would be great to include Gentoo within your test.
I'm pleasantly surprised too actually see so much difference between various OSes. I came in expecting them to be pretty even across the board.
*Debian:* They had us in the first round, not gonna lie.....
Is he seriously use debian 5 at the time debian 10 were already came out?
Why tho?
@@kerduslegend2644 Good question bro..
Boot speed was not fair, I think the desktop environments should be same on both the distros, i.e. the lightest Desktop available and should be compared to the lightest version of windows wiht no unecessary apps on startup. Good Effort overall, keep up the good work.
Moral of the video: Debian has the best performance.
@LinuxGeek Arch is meme not an os .
but when Debian sucked, it sucked big! Arch was more consistent during the test
@@linuxinside6188please elaborate
Hello TechHut, thank you very much for the great content, I have a question about the camera you are using for your RUclips channel? If you can share the details I will really appreciate it! Keep the great work!
Debian has always been the OS for me
Doesn't Debian ship the ESR version of Firefox? That might explain the performance difference in the web browsing bechmarks.
One thing is very clear: Arch > Debian
It depends
where's opensuse ?
Where's Manjaro
@@BittyDaProto manjaro is a subset of arch and it will always be slower than arch so why bother?
@@alenpaulvarghese Well... It Has a better gui
Please align the graphs in your result slides. Example @8:28. Debian won but it looks like Arch is shorter.
I don't like PizzaHut but I like TechHut
PizzaHut is awful!
very enjoy this series :D
Да уж, а теперь покажи нам как игры в Linux с обработкой шейдеров Vulkan работают в сравнении с Directx на винде.
Hello techHut, what is the desktop wallpapers used in the archlinux notebook?
Arch also has different kernel versions, one of them is Zen Kernel which is an optimized version supposed to be "the best" in terms of performance. You should try this on next.
The best in performance will always be the one tailored & built for your specific hardware and cpu, I use zen because of the features, I can't bother to test performance
The best now is TKG-PDS for single core, and TKD-BMQ for multi core, custom compiled for your system of course.
Hey TechHut, how about a challenge: Using Raspberry 0 w as your main PC for a day or three?
I'm sugesting this because in 2012, Chris from Explaining Computers used Rpi2 as his PC for a week. Now in 2021, covid made more jobs change to the online model, and here in Brazil the prices of computers got very high. As i'm using my Rpi4 for main computer for a whole year, ive had a really good experience on manjaro with sway etc. Many wonder what can be used or not in Rpi0w, so this test could be great. Some details to make the sugestion working:
-SYSTEM: raspberry OS. Maybe you get a better experience with arch, dietPi or even Gentoo idk.
-DESKTOP ENVIROMENT: change the LXDE to sway or i3 (novaspirit has a great tutorial on this). Maybe 540p resolution would be better
-PROGRAMS: abiword instead of libreoffice, gimp?, firefox with lightweight extensions or some terminal browser (i can sugest more tips on this topic)
-HARDWARE RECOMENDATIONS: SSD or HD with external powering instead of the microSD memory, a good microSD or even leepspvideo's PiSafe's eMMC to boot, keyboard with dongle, cable management, and of course a great usb hub or the GPIO shild with more USBs
-CHALLENGES: try typing some text, light image and audio editing, youtube playback with h264ify or pasting the link in vlc, gaming (even light PS1 games works), benchmarks, light versions of sites like gmail and facebook, video or audio conferences, multitasking in general with htop, wine, DRM, ssh, cmatrix of course, etc
I wonder if different linux desktops would have different results in 3D benchmark? I once did the Half-Life 2 Lost Coast benchmarking using ArcoLinux with different desktops and had interesting results.
I'd be curious to see the key libraries/package differences aside from the kernel that might be contributing to the differences in the debian and arch system. also, if there were any driver/cpu microcode version differences, and any differences in the default cpu power settings for the distro or if change the amd processor profile is different in different distros.
Shouldn't windows technically run with antivirus to more actually reflect real life scenarios?
There is Windows defender
good joke.
then linux should run one too
@@orkhepaj it doesn't have one on by default (cause it doesn't need one for 99.9% of the viruses out there)
@@atiedebee1020 cause nobody is using linux , no wonder it is
Even though Debian came dead last in all the testing, I still think it is an amazing OS. I have it installed on my VAIO.
2:14 - Windows launched it so slow because of the Windows Defender
I'm 99% sure.
Edit: I also found that Linux have loaded most of the libraries (they are needed for DE) while Windows does not (it uses Win32 API and not GTK+)
The biggest Speed gain is when actually using the OSs .. Any Linux Distro will smoke Windows .. The system lag with Windows is massive, application crashes, update hell and virus issues when using Windows. Open Source FTW !
I use manjaro xfce. That's it, I'm switching to ArchLinux
How’d it go ;)
it's effectively the same thing just with more curated packages, you'd get the same performance as plain arch with xfce.
what file system used on debian and on Arch?
good job, have a nice year.
OS basically doesn't affect hardware productivity in single running application. boot time isn't a productivity parameter
To be fair, Endeavor uses XFCE so it uses less resources than other OSes.
You ruined the whole video. Public service you have provided.
@@Jameshazlett As I said in a another comment "I've done other benchmarking videos with GNOME and XFCE distros. The difference is so minor in tests like this unless you're running a really low end system. The DE might add half a second to boot and opening the file manager may take a tenth of a second longer to open. I went with that what the distros default DE was. "
XFCE is the second-most bloated GUI, after GNOME.
Curious to see if Debian with updated Mesa would've scored more on the GPU test. That's a fun test you've done, thank you. Do you think Ubuntu would've been more appropriate? Asking because Debian's philosophy is focusing on rather outdated, albeit well-tested, packages and none of the other two OSs in the test are *that* stale in terms of package version.
FreeBSD and Debian always, don't have no time and bandwidth for Arch or any of the distros having lot of upgrades, even Ubuntu.
I´m agree with you, the best Linux distro imo based on years of experience its Debian. And on the top you have FreeBSD.
I use Manjaro for the same reason
@@CesarPeron Manjaro has limited updates? I doubt that, even Ubuntu has regular updates.
At home I have an old HP stream 13 netbook of 2015. It still works nicely with MX linux on emmc internal drive and also manjaro on a micro sd card. Of course I use light desktop environments such as lxde and xfce... With installed windows 7 I remember impossibility to watch full hd videos. Indeed on Linux with mpvplayer I can watch whatever. Great linux!! How do you score my English I posted from Italy?
Pick your OS (poison) for the need. Those that like Windows will never switch to Linux; either due to feeling it is too complicated or due to software compatibility.
As for Linux, I dual boot (separate drives) as I’m a Windows support / system admin and that is the corporate environment I need to keep up with, but I really like my Arch install. To me, Arch just flows so much better than Debian based systems.
Here again, there’s Debian fan boys that will scathe my comment; and Arch users that will praise me for running vanilla Arch rather than Endeavor or Manjaro.
Which is another reason why some won’t switch to Linux. The animosity and outright hostility towards Windows is one thing; distro to distro is just sad for the community at large.
ya, compatibility... speed is fine, but if i cant do what i want to whats the point....
first software i install are: key manager, XmouseButtonControl, hot virtual keyboard, directory opus, lightscreen, 7 stickyNotes, MT5/MLQ5 trading, potPlayer, notepad++, NetLimiter 4, Tixati, sordum topmost, need the magnifier, +++...
what am i gonna do with any Linux???... lol
Well, it's very interesting, but I'm not sure how would it turn out on completely different hardware. I mean - the speed of various tasks may strongly depend on the driver software. Some drivers might work faster on one OS and slower on another. So - on those laptops Windows is slower. But I'm not sure, maybe there are hardware configurations that would give completely different results. Anyway, thanks for the test, it was very interesting to watch. BTW, I remember the times the Linux struggled with 3D graphics. The times of very poor GPU drivers. I see it changed a lot.
I use debian by the way.
I use Gentoo and Arch.
Which branch Stable , testing or unstable ??
I have triple boot on my laptop. Windows to play games. Arch as Main OS. And Linux Mint for the emergency time when I f*** Arch up.
ur german, so of course you do!
An updated video would be really nice
You should have used pop os or ubuntu instead of debian, debian is amazing with stability but if you need speed is not the best choice..
Pop!_OS and Ubuntu are literately forks of Debian.
@@TechHut Yes, but packages on debian are really old and sometimes makes performance bad, ubuntu and pop are far superior for people who needs performance..especially for gaming.
wow a comparison we actually care about. I do wish there were more real world tests. opening browsers and other software that people use often.
Now you need to do all these test in a Garuda system with the Zen kernel.
On my PC Arch is faster than Debian but I love Debian more and it is my daily use oh and once the system is started the performance is the same.
Same,i prefer debian for being more casual
Hello from the future, and thanks for the good content
can you please do: void-linux vs Arch vs windows 10
Good idea
This reminded me of the Hennessy/Patterson textbook. A is better than B, B is better than A, or A/B are equals depending on the benchmark criteria.
It is not fair for you to use xfce in Endeavor and gnome in Debian. Even so, Debian is almost on par with Endeavor OS, even faster
I've done other benchmarking videos with GNOME and XFCE distros. The difference it's so minor in tests like this unless you're running a really low end system. The DE might add half a second to boot and opening the file manager may take a tenth of a second longer to open. I went with that what the distros default DE was. 🤓
@@TechHut does debian have a default de? I am using it and it just asks me which de to use during installation. How do you know that gnome is default for debian
Default is the wrong word. It's the first option and it's what they commonly use in demonstrations. The main point of what I said was the lack difference on systems with okay specs.
@@TechHut ok , i got it
@@cevizagaci31 In Debian, if you select "Debian desktop environment" during installation, and you don't actually select which environment below, then you will end up with GNOME anyway, so, technically, it's the default, though a very weak default since it asks you what you want.
This also brings up another point, though. I believe that Debian with GNOME uses Wayland by default, so that could be another difference in the general environment being used.
Can you tell us the power consumption.
If OS upgrading is also considered, then Arch/Debain >> windows
Windows runs smoothly in a fresh install, a year later you feel the need to buy a new computer, debian/arch runs smoothly in a fresh install and a year later runs better lol
@@izraell hahaha
@@izraell Never had that experience after 5 years.
@@Codyslx Probably because your Windows was slower since the start, lol.
@@kevinyoliveira68 How do you know that? You've never even seen my pc.
During Unigine Valley Benchmark run why does Windows report Build 9200 Windows 8? I thought this was a Windows 10 based test comparison.
No matter what,Daddy of all is the most beautiful,stable and the best Linux based distro ever - Debian .
i'm not so sure about this test. On my pc, i tested arch (pure) and debian , arch was way slower (both with gnome)
@@laurentiucalonfir Phoronix does many more tests and more or less Debian is in neck to neck with Arch. With wild variation of Arch as it has regular changes.
Booting into OS is NOT a benchmark. It is NOT a process. It's a random collection of custom services YOU DECIDE TO START.
Similar with "starting application". It's NOT a benchmark. Do you use caches? Prefetch? etc. etc. etc.
Slower might be better, and might be not. Depends.
"YOU DECIDE TO START" this was the base installation of all of these operating systems with nothing changed other than the login requirements for all of them.
I'm a Linux user but my i5 6th gen W10 machine boots from completely off to desktop in 7 seconds lol.
yeah win10 is pretty good
Win10 probably has fast startup on by default, which is hibernation, you can disable it to see how fast it really is.
@@user-dc9zo7ek5j it's a desktop. From the time I power on at the back of the psu, turn the machine on and sit down I am on the desktop. Also fast boot in the bios isn't active.
@@SleepingSavage Well then, 7s from totally turned off is really fast. My linux boots for 16 seconds, with all the bios logos and the grub delay. And I did not mean fast boot in bios, I meant fast startup, it's a special windows feature after win 8, that improves the startup by actually not shutting down fully.
@@user-dc9zo7ek5j yep I'm still surprised by how quick it is. Fast start up is not active. Hibernation states (of any kind) are useful on laptops but personally I never use them on anything. Just a personal preference.
Love your content man, great work. I would love to see a speed test between Ubuntu - Manjaro - MX Linux (KDE)
or duel > ReaL vs Madrid ;)
@@kafeshop Oh im sorry, i didnt realize that you are one of the dudes that seriously believes that debian and ubuntu are the same... Or even better that all linux distros are the same. Lol. Educate yourself and make your own research. ;) Its like you are saying that a VW Golf and a Seat Ibiza are the same car only cause the engine is common on both cars. Nope. :)
@@Alexandros_Alpha calm down PLz -> try arcolinux & satisfy your fashion needs ...
@@kafeshop Well, when you are a professional and not a hobbyist and you need something to work perfectly out of the box, you give some extra attention to the details of each distro. No no man, its not about fashion, seriously its obvious that you are using linux for fun and not for productivity.
@@Alexandros_Alpha its about procceses in unix /bsd ... and kernel tunning > srry ...
Please write it as endeavoros, not Arch
Could you please next time do vanilla arch and vanilla debian with the same DE for a fair comparison?
why is that fair?
@@orkhepaj because it compares how both base systems behave and which system is better for which activity and to mesure pure performance, as arch and debian are very different. The derived distributions modify which services and programs run in the background and desktop environments have an important impact on performance, thus makes the comparison very unfair and very difficult, and what adds to that is the fact that he used vanilla debian and not vanilla arch linux but a lightweight distribution based on it, it is like comparing a plane with a car. I think he should compare either the base systems, or a lightweight distributions of BOTH systems
@@nournemroud plane with a car?:O
nah these tests should be done on used and tailored systems , like what a normal person would have after a month or two
@@orkhepaj yeah they should be tested on regular hardware
so question is win or arch.
i say both :D
great vid btw.
But like Debian vs Arch
I think I have the same monitor as you, is that the Sceptre C30 Ultrawide?
I think so I have the 1080p model
Arch is love arch is life!
Please update this test, with Debian 11
@@kyzo_124 And -15% performance in Windows side :(
Windows for me
Never had much good experience with Linux in general..
Ubuntu was the best but still had a few issues like screen tearing even tho latest drivers were installed and vpn not working
Debian didn’t boot at all every time after installing 4 times in a row
OpenSuse was good.. but many key features were not working like simple video playback, suspend issues and again screen tearing
And for me it’s not worth to invest hours to fix this
Linux as Main OS is cool if you have the time to work everything out .. but for productivity it’s just not near reliable as Windows is
I know those experiences are subjective, just my opinion
Also one example: I created a windows installer on USB one year ago and this works no matter how often I use it
Making a USB installer for a Linux distro has some luck to it .. sometimes it works, sometimes it reports errors during installation
I don't know why but whenever I boot linux on a separate drive or an usb stick, something happens to windows and cpu usage goes down. For example, before I used linux I had 15 percent cpu usage in windows. After using linux, it went down to 7 percent. Now it is at 3 percent. That concerns me.
As for your post, linux works fine so does windows. I have no problem using both
You sir earned yourself a sub. I love your content.
now I'm just imagining "how fast XP was?"
XP will be close to Linux antiX
The difference is - antiX is supporting modern hardware and software while WinXP is not.
Hard to know if this is really apples to apples. If say you used snap on debian to install gimp you'd expect it to be somewhat slower than on arch.
Yeah Like Debian doesn't have gimp in its repo.