Jump to content
Sign in to follow this  
jarmstro

Installing Linux for X-Plane

Recommended Posts

1 hour ago, HumptyDumpty said:

Does my review count ?

I must have missed it! Can you post your results again real quick? (I always tell people to search/read the whole thread but here even I am too lazy to go through 25 pages 😉)

Share this post


Link to post
Share on other sites
9 hours ago, RXP said:

I'm pretty certain your PM box will get flooded with requests for results and this might just be easier posting them here 🙂

Like @Janov is saying, the purpose of this topic was to evaluate how much worth it really is in practice, and @mSparks must be commended for helping you ride along the way (and let us learn a few Linux tricks in the process too). You'd left most of use craving not having the end results after all these steps and discussion pages!

I've re done my driver install and thought it had made a difference but I think it was placebo. I've tried lots of scenarios and weather conditions.

I am getting a 10% boost in FPS in Linux over Windows. In use without an FPS counter on screen there's no difference I can notice.

- I seem to notice far more fluctuation in FPS in Windows.

-At high settings Linux is more jerky and stuttery.

-Linux loads faster.

 

  • Like 2

Share this post


Link to post
Share on other sites
1 minute ago, Janov said:

I must have missed it! Can you post your results again real quick? (I always tell people to search/read the whole thread but here even I am too lazy to go through 25 pages 😉)

My initial findings were 30% better in Linux using the same addons / aircraft / lua script.   was going to say 50% better :biggrin: 

Now this is my system config

AMD FX 4300

RX 550 2GB 

12GB DDR3

 So I fired up WIndows XP right now  , and hmm it seems to me that Linux is still ahead.  I will still trying different scenarios on both platforms.  

 The only thing I am not using in LInux is Enhanced CLoudscapes , which i am in windows , so that is one scenario I might not be able to test because I am not able to compile it correctly due to some error.    

I don't know if this my imagination but the same Lua script causes windows to tank the fps but remains stable in Linux.  

All the stuff in XP LInux is exactly the same from Windows and also the graphics settings.

 still very early to say if Linux holds the edge.  We need more tests like this to confirm.

  I was using XP in Linux extensively pre-vulkan due to the fact the AMD windows drivers for Opengl was locked which every AMD user will know , using Mesa was amazing in Linux pre-vulkan. But that is not a right review.

 

 

  • Like 2

Ryzen 5 1600x - 16GB DDR4 - RTX 3050 8GB - MSI Gaming Plus

Share this post


Link to post
Share on other sites
1 hour ago, HumptyDumpty said:

The only thing I am not using in LInux is Enhanced CLoudscapes , which i am in windows , so that is one scenario I might not be able to test because I am not able to compile it correctly due to some error. 

That's a pretty big factor TBH... I presume you benchmarked both without this addon enabled?

  • Like 2

Share this post


Link to post
Share on other sites
1 hour ago, tonywob said:

That's a pretty big factor TBH

And in addition one must be VERY careful making comparisons with clouds in the scene - often these are random and in one session you may see a lot of them, in the next session there is a big "gap" in the clouds ahead and framerate goes up. One should always test with NO clouds, and if clouds are to be part of the test, always use a solid overcast, not a "scattered" or "broken" setting.

  • Like 1

Share this post


Link to post
Share on other sites
1 hour ago, tonywob said:

That's a pretty big factor TBH... I presume you benchmarked both without this addon enabled?

Yes did with no clouds , but I can try it again and post the results. 

 

 


Ryzen 5 1600x - 16GB DDR4 - RTX 3050 8GB - MSI Gaming Plus

Share this post


Link to post
Share on other sites
18 minutes ago, Janov said:

And in addition one must be VERY careful making comparisons with clouds in the scene - often these are random and in one session you may see a lot of them, in the next session there is a big "gap" in the clouds ahead and framerate goes up. One should always test with NO clouds, and if clouds are to be part of the test, always use a solid overcast, not a "scattered" or "broken" setting.

did with no clouds too .  What i can do later is move EC also from WIndows and use CAVOK and post the results. 

 

  • Like 1

Ryzen 5 1600x - 16GB DDR4 - RTX 3050 8GB - MSI Gaming Plus

Share this post


Link to post
Share on other sites

EC is significantly faster at drawing clouds than default, especially overcast

except in certain weather situations they can complete wreck fps (which is why ece has a target gpu time).

Also

(and yall thought linux was complicated, that much effort to get linux performance on windows.... reaaaally?)

Edited by mSparks

AutoATC Developer

Share this post


Link to post
Share on other sites
20 minutes ago, mSparks said:

(and yall thought linux was complicated, that much effort to get linux performance on windows.... reaaaally?)

It is complicated,  because we have been using it since many years it's much easy for us. Even the most user friendly distros will seem much complicated .

Was off Linux for a very long time and got back into it because clients wanted squid / postfix and a few other stuff in house and was completely lost  ( the pandemic now has made them switch to a Cloud )   , still not so familiar anymore with Linux . 

The most now I use in Linux is the teminal to access the servers (yes i can use putty in windows )  , but now most of my setups are Pfsense based so it's a web gui because all are moving to a cloud and keeping pfsense just for office internet and protection is fine , only rarely do i need to access the terminal of Pfsense.

Also the 30% performance difference is hypothetical for now, more tests needed to be sure. 

Edited by HumptyDumpty

Ryzen 5 1600x - 16GB DDR4 - RTX 3050 8GB - MSI Gaming Plus

Share this post


Link to post
Share on other sites
10 minutes ago, HumptyDumpty said:

It is complicated

I also genuinely hate them. I'd be great if I could take what I know back into medieval times where the main thing to do with your time is hack each other to death with swords.

11 minutes ago, HumptyDumpty said:

Also the 30% performance difference is hypothetical for now, more tests needed to be sure.

The performance that can theoretically be gained is in two very specific areas.

Disk access, by the nature of modern linux filesystems being much more efficient and pumping around files and tuned for SSDs

Multithread performance - all the bits of XP and plugins that rely on doing things in parallel. Manifests as a smaller CPU time, and only as long as threading is the bottleneck to completing an entire frame (as opposed to, say, bus speed).

GPU time is more or less equal on both (in fact may even be marginally slower on Linux under very heavy GPU load, and ignoring AMD GPUs which are generally borken everywhere, but more so on Linux). 

 


AutoATC Developer

Share this post


Link to post
Share on other sites
7 minutes ago, mSparks said:

I also genuinely hate them. I'd be great if I could take what I know back into medieval times where the main thing to do with your time is hack each other to death with swords.

The performance that can theoretically be gained is in two very specific areas.

Disk access, by the nature of modern linux filesystems being much more efficient and pumping around files and tuned for SSDs

Multithread performance - all the bits of XP and plugins that rely on doing things in parallel. Manifests as a smaller CPU time, and only as long as threading is the bottleneck to completing an entire frame (as opposed to, say, bus speed).

GPU time is more or less equal on both (in fact may even be marginally slower on Linux under very heavy GPU load, and ignoring AMD GPUs which are generally borken everywhere, but more so on Linux). 

 

No idea about XP multithreading , the only thing about that I know is that Austin did mention the FM uses multithreads. 

If  EC works with Linux VUlkan it would tell us the  comparison to windows 


Ryzen 5 1600x - 16GB DDR4 - RTX 3050 8GB - MSI Gaming Plus

Share this post


Link to post
Share on other sites

Just tested :

NO ENHANCED CLOUDSCAPES IN WINDOWS / LINUX.    Scenery / Aircraft / Plugins / Lua exactly the same / Graphics settings exactly the same except in LINUX have enabled Shadows on SCenery.    Both using VUlkan.

In WinDOWS :

ICAO : FACT  using real Wx  

Fps on runway : 20-21  

Turn Towards Table Mountain : drops to 18-19 max 20. Outside the helicopter using "c" jumps to 31-35.

NO LIVETRAFFIC PLUGIN IN WINDOWS 

In Linux :

ICAO : FACT using real Wx.

Fps on runway : 25-27 , Outside the helicopter using "c"   fps jumps to 40ish,.

Turn Towards Table Mountain : 

 This is what I am seeing. 

LIVETRAFFIC PLUGIN ENABLED. 

 

Suggest some more testers @jarmstro  any update ? 

 

  • Like 1

Ryzen 5 1600x - 16GB DDR4 - RTX 3050 8GB - MSI Gaming Plus

Share this post


Link to post
Share on other sites
16 minutes ago, HumptyDumpty said:

Just tested :

NO ENHANCED CLOUDSCAPES IN WINDOWS / LINUX.    Scenery / Aircraft / Plugins / Lua exactly the same / Graphics settings exactly the same except in LINUX have enabled Shadows on SCenery.    Both using VUlkan.

In WinDOWS :

ICAO : FACT  using real Wx  

Fps on runway : 20-21  

Turn Towards Table Mountain : drops to 18-19 max 20. Outside the helicopter using "c" jumps to 31-35.

NO LIVETRAFFIC PLUGIN IN WINDOWS 

In Linux :

ICAO : FACT using real Wx.

Fps on runway : 25-27 , Outside the helicopter using "c"   fps jumps to 40ish,.

Turn Towards Table Mountain : 

 This is what I am seeing. 

LIVETRAFFIC PLUGIN ENABLED. 

 

Suggest some more testers @jarmstro  any update ? 

 

As I said above I get about a 10% increase over windows. ( A generalisation but it's about right. It's in that region.) Which is not to be sneezed at. I'm guessing some systems will gain more than others.

And there's no stopping me now. I've installed Wine and Lutris and am about to try out DCS. 😀

Edited by jarmstro
  • Like 1

Share this post


Link to post
Share on other sites
7 minutes ago, jarmstro said:

As I said above I get about a 10% increase over windows. ( A generalisation but it's about right. It's in that region.) Which is not to be sneezed at. I'm guessing some systems will gain more than others.

And there's no stopping me now. I've installed Wine and Lutris and am about to try out DCS. 😀

LOL. All the best , the only one that ran OOB for me is il2 , but the fps is half of windows.  DCS refuses to run , Have Arma 3 download but not tried. 

around 20% gain is what I am seeing in linux. If i enable shadows in windows it will go below 10 but in linux most places i am seeing a flyable response.

It's also ahead at Vegas / LA , I am using the US West Life scenery for it and Orthos. 


Ryzen 5 1600x - 16GB DDR4 - RTX 3050 8GB - MSI Gaming Plus

Share this post


Link to post
Share on other sites
20 minutes ago, HumptyDumpty said:

LOL. All the best , the only one that ran OOB for me is il2 , but the fps is half of windows.  DCS refuses to run , Have Arma 3 download but not tried. 

around 20% gain is what I am seeing in linux. If i enable shadows in windows it will go below 10 but in linux most places i am seeing a flyable response.

It's also ahead at Vegas / LA , I am using the US West Life scenery for it and Orthos. 

The setting I used are:

No. of World Objects - Max

Reflections - One notch up

Visual Effects - One notch down

Textures - One notch down

AA - In the middle

FA - In the middle 

Shadows - on

Vulkan - on

OrbX and posh airports. Clear sky's. Default Cessna.

Example:

Flying around EGGD I'm getting around 50fps Linux and 45fps Windows. The most I saw was actually in Windows - 62fps

 

Edited by jarmstro

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this  

  • Tom Allensworth,
    Founder of AVSIM Online


  • Flight Simulation's Premier Resource!

    AVSIM is a free service to the flight simulation community. AVSIM is staffed completely by volunteers and all funds donated to AVSIM go directly back to supporting the community. Your donation here helps to pay our bandwidth costs, emergency funding, and other general costs that crop up from time to time. Thank you for your support!

    Click here for more information and to see all donations year to date.
×
×
  • Create New...