Update 4090 results.

Author: geerlingguy

_cards_gpu/gigabyte-nvidia-rtx-4090.md

@@ -3,73 +3,19 @@ layout: card
 title: "Gigabyte Nvidia RTX 4090 OC"
 picture: "/images/gpu-nvidia-rtx-4090-oc.jpg"
 functionality_cm4: "None"
-functionality_pi5: "Untested"
+functionality_pi5: "Partial"
 driver_required: "Yes"
 github_issue: "https://github.com/geerlingguy/raspberry-pi-pcie-devices/issues/504"
 buy_link: https://amzn.to/3KaTN26
 videos:
   - https://www.youtube.com/watch?v=l9dItRUjQ0k
 ---
-The RTX 4090 is supported by both Nvidia's proprietary driver and the open source Nouveau driver in the Linux Kernel.
+The Quadro RTX 3060 is supported by both Nvidia's proprietary driver and the open source Nouveau driver in the Linux Kernel.
 
-It will require a 1x to 16x adapter, or some other means of adapting the 16x-sized card pins into the 1x slot on the CM4 IO Board. You also need to supply at least 450W through the card's ATX12VO power plug.
+You need to provide a physical x16 slot, and supplemental PCIe 12-pin power to give the card power (it needs a lot...).
 
-There are two ways to try installing the driver:
+## Driver
 
-### Proprietary Nvidia driver
+Currently the easiest way to get this card working with a Pi 5 is with [mariobalanica's patch to the Nvidia open GPU kernel module driver](https://www.jeffgeerling.com/blog/2025/nvidia-graphics-cards-work-on-pi-5-and-rockchip).
 
-After flashing the 64-bit Pi OS beta to my Pi, I made sure to upgrade everything on it, and install the kernel source, so the Nvidia driver would compile:
-
-```
-sudo apt-get update
-sudo apt-get -y dist-upgrade
-sudo apt-get install -y raspberrypi-kernel-headers
-sudo reboot
-```
-
-After rebooting, if you have X server running (e.g. you are logged in via GUI), you need to log out. If you're running this over SSH or serial connection, you can quit X server with: `sudo systemctl stop lightdm`.
-
-Now, download Nvidia's latest [AARCH64 Driver for ARM 64-bit processors](https://www.nvidia.com/en-us/drivers/unix/linux-aarch64-archive/), make the downloaded `.run` file executable, and run it with `sudo`:
-
-```
-chmod +x NVIDIA-Linux-aarch64-525.89.02.run
-sudo ./NVIDIA-Linux-aarch64-525.89.02.run
-```
-
-After doing that, I was unable to get any display outputs through the card, and when running `startx`, I would get the error:
-
-```
-...
-[   130.737] (==) NVIDIA(0): Using gamma correction (1.0, 1.0, 1.0)
-[   130.737] (EE) 
-[   130.737] (EE) Backtrace:
-[   130.739] (EE) 0: /usr/lib/xorg/Xorg (OsLookupColor+0x188) [0x555bacc538]
-[   130.740] (EE) unw_get_proc_info failed: no unwind info found [-10]
-[   130.740] (EE) 
-[   130.740] (EE) Segmentation fault at address 0x124
-```
-
-### Nouveau (open source) driver
-
-To get the Nouveau driver loaded, you have to [recompile the Linux kernel for Pi OS](https://github.com/geerlingguy/raspberry-pi-pcie-devices/tree/master/extras/cross-compile).
-
-Go to `Device Drivers` > `Graphics support` on the `menuconfig` step, and select the `Nouveau` driver to install.
-
-If you want to be able to boot the Pi all the way, make sure to blacklist the `nouveau` kernel module:
-
-```
-sudo nano /etc/modprobe.d/blacklist-nouveau.conf
-
-# Put this inside the file and save it:
-blacklist nouveau
-```
-
-Now, after a reboot, you can explicitly load the module and see what happens by following `dmesg` in another terminal session:
-
-```
-sudo modprobe nouveau
-```
-
-TODO: What happens?
-
-See the linked GitHub issue for more details.
+See the linked GitHub issue for more information.