1. Home
  2. Linux
  3. Overclock nvidia graphics cards linux

How to overclock Nvidia graphics cards on Linux

Nvidia graphics cards are a standard on Linux because of the companies good proprietary graphics drivers. Out of all the other GPU providers, Nvidia provides decent (albeit closed-source) graphical drivers that allow video games to run pretty well on the platform.

Still, Nvidia GPUs sometimes under-perform on Linux compared to Windows, due to the underlying technology of Linux. Thankfully, it’s easy to get around these performance downsides by overclocking the hardware.

There are a few ways to overclock Nvidia graphics cards on Linux, but the easiest way to do it with little trouble is by installing the GreenWithEnvy graphical tool, as it allows users to use an easy GUI to provide info and overclock the Nvidia graphics cards.

WARNING: do not try to overclock your Nvidia graphics card unless you’re confident you have a power supply that can handle the extra overhead. To learn about your power consumption, and what the stock power draw is, list out your components in this PSU calculator.

Install GreenWithEnvy

There are a few ways to install the GreenWithEnvy Nvidia overclocking application on Linux. According to the developer, the preferred way to get the app working is through Flatpak. The reason for Flatpak is that the software comes ready to go, with no need to compile anything.

Before we go over how to install the GreenWithEnvy application in this way, it’s important to talk about the Flatpak runtime, because without it, installing and running this program isn’t possible.

The Flatpak runtime is quite easy to install on a majority of popular Linux distributions. To get it going, open up a terminal window. Then, click on this guide here to learn about how to get everything working.

After the Flatpak runtime is done installing, it’s safe to install the GreenWithEnvy application on your Linux system by entering the following commands in a terminal window.

flatpak remote-add --if-not-exists flathub https://flathub.org/repo/flathub.flatpakrepo
flatpak install flathub com.leinardi.gwe

Other ways to install GreenWithEnvy

Can’t use Flatpak and need to use the GreenWithEnvy application on your system? Lucky for you, the developer makes it pretty easy to install the program on a wide variety of Linux distributions by compiling it from source.

Compiling programs from source on Linux starts by installing all of the dependency files needed to build the software without any errors. To get these dependencies working, launch a terminal window and enter the commands that correspond with the OS you are currently using.

Ubuntu

sudo apt install git meson python3-pip libcairo2-dev libgirepository1.0-dev libglib2.0-dev libdazzle-1.0-dev gir1.2-gtksource-3.0 gir1.2-appindicator3-0.1 python3-gi-cairo appstream-util

Debian

sudo apt-get install git meson python3-pip libcairo2-dev libgirepository1.0-dev libglib2.0-dev libdazzle-1.0-dev gir1.2-gtksource-3.0 gir1.2-appindicator3-0.1 python3-gi-cairo appstream-util

Arch Linux

On Arch Linux, there’s no need to deal with installing dependencies manually for GreenWithEnvy. Instead, build and install the AUR package in a terminal with the commands below.

sudo pacman -S git base-devel
git clone https://github.com/trizen/trizencd trizenmakepkg -sritrizen -S gwe

Fedora

sudo dnf install desktop-file-utils git gobject-introspection-devel gtk3-devel libappstream-glib libdazzle libnotify meson python3-cairocffi python3-devel python3-pip redhat-rpm-config

OpenSUSE

Sadly, the developer doesn’t outline dependencies to install for OpenSUSE. Try using the Fedora dependencies above, as they are similar. Or, if those don’t work, stick to the Flatpak release.

With the dependencies taken care of, build and install GreenWithEnvy from the source code using the following terminal commands.

git clone --recurse-submodules -j4 https://gitlab.com/leinardi/gwe.git
cd gwe
git checkout release
pip3 install -r requirements.txt
meson . build --prefix /usr
ninja -v -C build
ninja -v -C build install

Finally, install the Nvidia-xconfig package on your system and then use it to set your GPU into overclocking mode.

Ubuntu

Assuming you’re using the latest Nvidia GPU drivers on Ubuntu, the nvidia-xconfig app should be installed.

Debian

sudo apt-get install nvidia-xconfig

Arch Linux

The nvidia-xconfig package comes with the default Nvidia drivers on Arch, so no need to install it as a separate package.

Fedora

sudo dnf install nvidia-xconfig

OpenSUSE

The nvidia-xconfig package is included with the OpenSUSE Nvidia drivers.

sudo nvidia-xconfig --cool-bits=12

Overclock with GreenWithEnvy

Overclocking a graphics card is a tricky thing. Generally speaking, it’s best to do tweaks in small increments. For best results, follow along with the guide and do not deviate unless you’re an expert!

To start the overclocking, launch GreenWithEnvy on your Linux PC. Once the application is open, take a look at the temperature of the card. Watch the stats for a bit to make sure that it doesn’t pass the “SLOWDOWN TEMP” or the “SHUTDOWN TEMP.”

Assuming the temperatures are good, move down to the “Power” section. Then, use the mouse to drag the slider up by about 5%, and click the “apply” button. For example, if your power is set at 120, move it to 125.

Use your computer for a bit and do something graphically intensive like watching an HD video or run a video game and assess if everything is running stable. If no problems persist, feel free to increase the power by another 5-10%.

Revert the overclock

Unhappy with your overclock? Open up the GreenWithEnvy application on your Linux desktop and locate the “Power” section. After that, look to the slider and set it back to default.

Note: unsure what the default is? It’s the vertical line on the slider.

After setting the power back to default on the slider, click “apply” to confirm changes.

2 Comments

  1. I run Kubuntu 22.04
    Ok, it says to run
    sudo apt install git meson python3-pip libcairo2-dev libgirepository1.0-dev libglib2.0-dev libdazzle-1.0-dev gir1.2-gtksource-3.0 gir1.2-appindicator3-0.1 python3-gi-cairo appstream-util

    and then what?

  2. Erm this was far too brief for example despite setting coolbits to 12 i cant move the power slider only thing i can change is gpu and memory offset I have 2070 mobile ti graphis card