Which GPU am I using?



This may be a weird question, but I seriously do not know the optimal config for my system. I have a laptop with an MX150 GPU from nvidia and obviously with built in graphics. As far as i can recall i installed manjaro using the non-free drivers for the card, since i only got a completely black screen when using the free drivers. Now I am unsure whether my GPU is actually being used and which driver config is suitable.

Here is some system info:

$ glxinfo | grep OpenGL
OpenGL vendor string: Intel Open Source Technology Center
OpenGL renderer string: Mesa DRI Intel(R) HD Graphics (Whiskey Lake 3x8 GT2) 
OpenGL core profile version string: 4.5 (Core Profile) Mesa 18.3.3
OpenGL core profile shading language version string: 4.50
OpenGL core profile context flags: (none)
OpenGL core profile profile mask: core profile
OpenGL core profile extensions:
OpenGL version string: 3.0 Mesa 18.3.3
OpenGL shading language version string: 1.30
OpenGL context flags: (none)
OpenGL extensions:
OpenGL ES profile version string: OpenGL ES 3.2 Mesa 18.3.3
OpenGL ES profile shading language version string: OpenGL ES GLSL ES 3.20
OpenGL ES profile extensions
$ inxi -G
Graphics:  Device-1: Intel UHD Graphics 620 driver: i915 v: kernel 
           Device-2: NVIDIA GP108M [GeForce MX150] driver: N/A 
           Display: x11 server: X.Org 1.20.3 driver: intel resolution: 1920x1080~60Hz 
           OpenGL: renderer: Mesa DRI Intel HD Graphics (Whiskey Lake 3x8 GT2) v: 4.5 Mesa 18.3.3
$ neofetch
CPU: Intel i7-8565U (8) @ 4.600GHz 
GPU: Intel UHD Graphics 620 

The arch wiki is somewhat confusing for me concerning this topic. It lists many options for “switchable graphics” (bumblebee, primus, nvidia-xrun, …). Do i even have what is considered “switchable graphis”? Basically what I would like to achieve is windows-like behaviour, i.e. the dedicated card should be used whenever applicable. Are there even any power saving improvements when completely disabling the dedicated GPU compared to it idling? What is my MX150 currently doing? Is it idling, does it have its drivers even installed? This situation is confusing me. What would be the best configuration to always enable the dedicated card?

Many Thanks.


setting up switchable graphics is not my bailiwick but i can answer the question in your subject line:

inxi reports both cards, but it tells you which one is currently running xwindows.


Alright, thanks. So my dedicated GPU is not at all in use? As I said, I dont necessarily need functionality to switch, although it might be interesting to see how big the power saving actually is.


I have similar configuration mx150 + integrated. As I understood to use dedicated card you should run

optirun <name_of_application>


Well this is the thing… If i used the card for gaming this might be fine, however this is a workstation machine. Many things I do have CPU and GPU “backends” (im not sure whether this is the correct term), and I want to make sure that the GPU is properly utilized. Take tensorflow as an example, it uses the GPU when it has one available to use implicitly. Would i have to optirun my python interpreter? Or how would that work? I assume that there are many applications that offload some work to the GPU if possible. I have also done a lot of graphics programming (OpenGL and OpenCL) previously, I would prefer that i could also do that on this machine using my manjaro config.
Further, the arch wiki actually discourages the use of bumblebee alltogether, so I would prefer a different solution…

Edit: So there is, apart from the explicit pages about bumblebee etc., also a page in the Optimus techology in general on the wiki. It mentions that the proprietary driver basically does in fact implement this technology also on linux. Does this mean that this would be the equivalent to the windows experience? Then the question still remains how big the benefit of actually disabling the card versus using the optimus technology really is… Anyways i guess I will simply go with the proprietary driver then. Still curious why this isnt the default behaviour when installing with the non-free driver option…


Um I cant speak for everybody,
The whole idea behind the laptop = battery power and portable and not games or anything 3D or on and on …

So what do you propose which would be better for laptop, or workstation or OpenGL programer?

Are you trying to use a hybrid motherboard for a workstation or are you using laptop for workstation?
If you are only ever connected to electric socket then for sure just use the Nvidia proprietary.


Not sure what exactly you mean by hybrid motherboard vs. laptop but this is entirely off topic. Im definitely not always connected to power, but obviously Im willing to sacrifice some battery life for proper GPU usage. I will try the proprietary drivers and see whether i notice any difference in battery life.