Hello Fedora People !! ;
Yesterday night finally after our crash and failing We succeed to install Nvidia CUDA driver in our Fedora 20 and also attach with optimus and only work when I active "optirun" with app and done.When I tell to method "how to" you can even use If you even single nvidia card just don't do optimus and use usual driver from "RPMFusion" and add to CUDA and It will work.
So let's start to install Nvidia CUDA to our Nvidia card.
If you wanna use Optimus + CUDA first things you gonna need is repository and also installing commands ;
Installing Packages in order to have Nvidia Optimus (a.k.a Bumblebee)
Then "reboot" for have running optimus.
CUDA Part ( CAN WORK EVEN SINGLE GPU CARD BASED PC/LAPTOP ) ;
Go to https://developer.nvidia.com/cuda-downloads website and download "Fedora 18" "RUN" files not "RPM" one because we gonna compile ourself and add to system in order to make it work.
Add Cuda Libraries in our system as a "$PATH" and test with "nvcc"
## Result ##
Another Test With Pyrit (Added CUDA add-on from source)
Another Looks From Blender and Activated CUDA ;
I wanna say thank you to "Doğan ÇİĞNAKLI" for his effort and help and encourage me to do that. Because of him I found the "compiler solution" and Nvidia CUDA based application enviroment set-up.
He also using "Fedora 20" and GPU Nvidia GT 740M GPU also working properly.
Every solution cannot be perfect but this is our solution I hope it will help to any readers who has "Intel/Nvidia" or Single Nvidia GPU card."
Source I used :
Doğan's code parameters :) (-override compiler our best one ) Like a our "startx"