Polymathic Perfraction: Nvidia

A completely random quote:



"The first principles of the universe are atoms and empty space; everything else is merely thought to exist." - Democritus


In this video, I'll be giving a general overview for:


- How to install Nvidia drivers for Linux Distros running the Zen Kernel, when an AUR package is not yet available.
- How to *approach* modifying an Nvidia AUR packages or create a new one.
- How to search the Nvidia Linux driver installer source to ensure you've updated the Linux patches to match the driver's expanded capabilities (though my knowledge here is woefully lacking).

- How to configure pacman to ignore the nvidia-* packages and still get it to build a DKMS to rebuild the modules when updating the OS.

- How a pacman hook can be configured to build the DKMS module.


I won't be carrying any of this out on the stream, but instead describing how the pieces fit together.


Garuda Linux runs the Zen Kernel, which has allow me to see unbelievable performance gains on my 2013 fully loaded Macbook Pro as well as my 2011 machine. It is seriously the difference between being able to open like 7 tabs in a browser to opening 100 tabs ... before seeing impacts on performance. Seriously, the differences were mind-blowing to me.


When I first ran Garuda, it was the first time I had seen OBS run the NVEnc encoders. me, NVEnc is the difference between being able to stream with ZERO dropped frames and 5% dropped frames... I made the mistake of trying to build `obs-studio-wayland` ... since then, I have spent ONE HUNDRED PLUS HOURS trying to get that NVEnc shit to work again. I even went to the lengths of beginning to cargo cult my own AUR Nvidia package. Though I have little to show for it ... but I've learned so much along the way (and knowing is half the battle)


If the stream quality is lacking, just keep watching it son, cuz I got 99 solutions but a fresh Mac ain't one.

My last programming stream was a bit of a bummer for me. Since I was still learning Julia, it was just embarrassingly bad, which is a shame because I have a great learning process for new languages, frameworks and pretty much everything.