guys please fedora uses my CPU instead of my Geforce GTX 1050ti.(see screenshot) How can I change it?
Do you have the Nvidia Drivers installed? If you want to use RPM Fusion for Nvidia, just run these commands to set it up:
sudo dnf update -y # and reboot if you are not on the latest kernel
sudo dnf install akmod-nvidia # rhel/centos users can use kmod-nvidia instead
sudo dnf install xorg-x11-drv-nvidia-cuda #optional for cuda/nvdec/nvenc support
Now, you might want to check out the RPM Fusion Optimus page to learn more about how to run things on the dGPU.
I just looked at my own specs and saw “AMD Turks” under graphics. I was puzzled!
Had to do some reading up to see it’s my old AMD 6670 graphics card that Fedora detected easily.
Good to know Fedora works well with old AMD A10-5800 APU and AMD graphics card. It runs smoothly too!
  In Gnome, clicking an icon with the Right Mouse Button should pop-up option “Launch using dedicated Graphics” or something alike, for nouveau
too (possible giving an enough boost to performance, without the need for using p-blob).
  Also there was some environment variable for this.
Hi @theclemi and welcome to Ask Fedora!
I have to say that if you just need stable and reliable desktop system – if you won’t play games on your Fedora system or use some other stuff requiring heavy 3D graphics – then in general you’ll be better off using integrated Intel graphics. It’s more stable and has fewer issues, if any.
As far as I know, the desktop itself won’t benefit at all from installing proprietary NVidia’s driver – and you can run into some issues with that.
As far as I know, Wayland – next generation graphics server Fedora uses by default – is still disabled by default when you install proprietary NVidia’s driver. NVidia’s driver always had issues with it, and I’m not sure if it’s resolved or not.
That said, you can certainly try proprietary driver – to see for yourself if you get any benefits from using it. It works well, I’ve used it in the past (but now I use Intel’s graphics or built-in nouveau driver on my Fedora boxes), and the issues I’ve mentioned are easy to resolve (are widely discussed and easy to find).
One more note, if you use RPMFusion as @cheeseeboi suggested – as you absolutely should if you would do it at all – then the proprietary driver is easy to remove afterwards. So you can try it, and then revert to default state if you don’t see any benefits.
it’s a good solution to use Nvidia GPU for fedora, simple and fast!!
https://www.if-not-true-then-false.com/2015/fedora-nvidia-guide
My experience with Nvidia is contrary. Any one who wants to have less headaches running an OS that is built on open principles should be careful running proprietary code on top of it. I had to revert to AMD GPU’s to escape the Nvidia pain while running Fedora.
I am probably missing out on GPU performance that tops the charts, but I am certainly enjoying less frequent troubleshooting on my Fedora box as a result. Time is of the essense. If it works well with the opensource ecosystem, I am game.
@simmon, I have to disagree!
It was discussed here on Ask Fedora quite a few times.
Using RpmFusion’s packaged for Fedora version is much better than using NVidia’s own universal run file, as that site suggests.
The driver itself is the same, built from the same source. But again, RpmFusion’s version packaged properly – meaning that:
-
it can be removed without leaving any leftovers on your system. Nvidia’s version known for doing so – and messing up future installations.
-
it will be updated automatically. Nvidia’s version won’t.
-
it can have some Fedora-specific patches if such are needed. I don’t think it does at all times (this can be easily be checked by inspecting the spec file) – but if some tweaks are needed, then package maintainer will include them, and user won’t have to search the Internet for some obscure errors and figure out how to deal with them.
It’s much simpler and cleaner to install a known good packaged version. And it’s a much better advice – especially for newer users.
I know. However, I would like to point out that there are other effective ways that can be applied immediately to practical use.it’s another way!
Nevertheless, there remains a part of the temporal problem of incomplete support. I can’t say that the manufacturer driver support is bad for a way to fix this, and I think the manufacturer needs more open source disclosure and active Linux support.
God bless…take care…everyone!