Getting my Quadro T1000 card to play nicely with RHEL 8.3

GPU Unix Graphics Linux
#1

I have an NVIDIA Quadro T1000 card on a system running a fully patched RHEL 8.3 running on the hardware (not as a VM guest OS) My hardware is a laptop (Dell 7750) @3840x2160 and two external monitors on a dock @1920x1080. I want to scale the builtin monitor to 200% and keep the extrnal monitors @ 100%. Red Hat tells me that their configuration utility does not “play nicely” with NVIDIA, so I need to use theNVIDIA configuration utility.

I’ve looked at nvidia-settings GUI, but idtdoes not seem to provide the needed functionality. Please advise.

#2

Doesn’t the normal monitor config in Gnome control center work?

#3

Red Hat discourages using GNOME’s monitor config with NVIDIA.

#4

Start nvidia-settings, go to
“X Server Display Configuration”, select monitor, hit “Advanced”. Enter different “ViewPortIn” dimensions, hit “Apply”
Though I would discourage doing that. First of all, the setting is not sticky, you would have to create an autostart item to do this on login. Second, you’d be working against Gnome’s built-in monitor manager which is likely getting confused when modes are set from the outside. Third, since you’re on a notebook you likely have hybrid graphics, so you will see a “Prime display” which is controlled by the igpu so the gnome control center has to be used.

#5

Thanks for your insights. I’ll pass them on to the RHEL engineers.

#6

Here’s the explanation from Red Hat:

Greetings,

Nvidia's package clobbers our graphics libraries, which breaks our graphics drivers. If you have a hybrid graphics equipped system, the intel or AMD built-in GPU (depending on your CPU) must be disabled at the BIOS level for the Nvidia GPU to work properly with their proprietary driver. Otherwise, most systems with hybrid graphics default to the on-die GPU to conserve power. With 3D acceleration effectively broken for our drivers when Nvidia clobbers our graphics libraries, you're left with no choice but to disable the on-die GPU in favor of the discreet Nvidia GPU.

Thanks and have a great day!
Best Regards,

Joe Wright, RHCE RHCVA
Software Maintenance Engineer, Desktop/Tools Team
Customer Experience & Engagement - North America
Red Hat, Inc
#7

Interesting answer, sounds a bit like fetched from the way-back machine.
How did you install the nvidia driver?

#8

the driver was downloaded from NVIDIA and installed using NVIDIA instructions.

#9

Ok, please don’t use a .run installer on Linux unless you really know what you’doing or just like to experiment.
RHEL provides their own nvidia driver repo
https://developer.nvidia.com/blog/streamlining-nvidia-driver-deployment-on-rhel-8-with-modularity-streams/
otherwise, you can e.g. use the rpmfusion repo
https://rpmfusion.org/Howto/NVIDIA

#10

Thanks. I’ve passed this on to my RHEL tech support guys (although you would think they already know about this).