All Products
Search

This Product

    Elastic GPU Service:Manually install a Tesla driver on a GPU-accelerated compute-optimized instance (Linux)

    Document Center

    Elastic GPU Service:Manually install a Tesla driver on a GPU-accelerated compute-optimized instance (Linux)

    Last Updated:Feb 02, 2026

    A GPU needs a Tesla driver to deliver high-performance computing (HPC) for deep learning and AI. The driver is also needed for smooth graphics in scenarios such as OpenGL, Direct3D, and cloud gaming. If you did not install a Tesla driver when you created a GPU-accelerated compute-optimized instance that runs Linux, you must install the driver manually. This topic describes how to manually install a Tesla driver on a GPU-accelerated compute-optimized instance that runs Linux.

    Procedure

    This procedure applies to all GPU-accelerated compute-optimized instances that run Linux. For more information, see GPU-accelerated compute-optimized (gn/ebm/scc series). You must install a Tesla driver for Linux that is compatible with the operating system of your instance.

    Step 1: Download the NVIDIA Tesla driver

    1. Visit the NVIDIA Driver Download Page.

      Note

      For more information about how to install and configure NVIDIA drivers, see the NVIDIA Driver Installation Quickstart Guide.

    2. Set the search criteria and click Search to find a suitable driver.

      Tesla驱动.jpg

      The following table describes the settings.

      Setting

      Description

      Example

      • Product Type

      • Product Series

      • Product Family

      Select the product type, series, and family that correspond to the GPU of your instance type.

      Note

      For more information about how to view details of a GPU-accelerated instance, such as the instance ID, instance type, and operating system, see View instance information.

      • Data Center / Tesla

      • A-Series

      • NVIDIA A10

      Operating system

      Select the Linux operating system version that corresponds to the image used by the instance.

      Linux 64-bit

      CUDA Toolkit version

      Select the CUDA Toolkit version.

      11.4

      Language

      Select the language for the driver.

      Chinese (Simplified)

      GPU information, supported driver versions, and CUDA versions for some GPU-accelerated compute-optimized instance types

      Item

      gn8v

      gn8is

      gn7e

      gn7i

      gn7

      gn6e

      gn6i

      gn6v

      gn5i

      gn5

      Product Type

      Data Center / Tesla

      Data Center / Tesla

      Data Center / Tesla

      Data Center / Tesla

      Data Center / Tesla

      Data Center / Tesla

      Data Center / Tesla

      Data Center / Tesla

      Data Center / Tesla

      Data Center / Tesla

      Product Series

      H-Series

      L-Series

      A-Series

      A-Series

      A-Series

      V-Series

      T-Series

      V-Series

      P-Series

      P-Series

      Recommended Tesla driver version

      570.133.20 or later

      450.80.02 or later

      460.73.01 or later

      450.80.02 or later

      410.79 or later

      Recommended CUDA Toolkit version

      CUDA Toolkit 12.4 Update 1

      CUDA Toolkit 11.0 Update 1

      CUDA Toolkit 11.2

      CUDA Toolkit 11.0 Update 1

      CUDA Toolkit 10.1 Update 2

      Note

      • This table lists GPU information for only some common GPU-accelerated compute-optimized instance types. Instances with the same GPU card share the same GPU information, such as product type, product series, and product family. For example, ebmgn7i and gn7i instances both use NVIDIA A10 GPUs, so they have the same product type, series, and family.

      • When you manually install a Tesla driver and a CUDA package, you must ensure that the driver version is compatible with the CUDA package version. For more information, see CUDA Compatibility.

    3. On the driver search results page, click Expand to view more versions.

    4. Find the driver to download and click View next to it.

      For example, select Data Center Driver for Linux x64 with driver version 470.161.03 and CUDA Toolkit version 11.4.

    5. On the product page for the driver, right-click Download and select Copy link address.

      驱动下载.jpg

    6. Remotely connect to the GPU-accelerated instance that runs Linux.

      For more information, see Log on to a Linux instance using Workbench.

    7. Run the following command to download the driver installation package.

      The driver download address in the command example is the download link from Step 5.

      wget https://us.download.nvidia.com/tesla/470.161.03/NVIDIA-Linux-x86_64-470.161.03.run

    Step 2: Install the NVIDIA Tesla driver

    The method for installing the Tesla driver varies depending on the operating system. The following sections describe the procedures.

    CentOS

    1. Run the following command to check whether the kernel-devel and kernel-headers packages are installed on the GPU-accelerated instance.

      sudo rpm  -qa | grep $(uname -r)

      • If a response similar to the following example is returned, the packages are installed. The response includes the version information for the kernel-devel and kernel-headers packages.

        kernel-3.10.0-1062.18.1.el7.x86_64
kernel-devel-3.10.0-1062.18.1.el7.x86_64
kernel-headers-3.10.0-1062.18.1.el7.x86_64

      • If the response does not contain kernel-devel-* and kernel-headers-*, you must download and install the kernel-devel and kernel-headers packages that match your kernel version.

        Important

        If the kernel-devel version does not match the kernel version, a compilation error occurs when you install the driver. Check the version number of kernel-* in the response and download the matching version of kernel-devel. In the sample response, the kernel version is 3.10.0-1062.18.1.el7.x86_64.

    2. Grant permissions and install the Tesla driver.

      For a driver for Linux 64-bit, use the .run file, such as NVIDIA-Linux-x86_64-xxxx.run. Run the following commands to grant permissions and install the Tesla driver.

      Note

      If you are using a Tesla driver in another format, such as .deb or .rpm, see the NVIDIA CUDA Installation Guide for Linux for installation instructions.

      sudo chmod +x NVIDIA-Linux-x86_64-xxxx.run
      sudo sh NVIDIA-Linux-x86_64-xxxx.run

    3. Run the following command to check whether the Tesla driver is installed.

      nvidia-smi

      A response similar to the following example indicates that the Tesla driver is installed.

      驱动版本.jpg

    4. (Optional) Enable Persistence-M mode using the NVIDIA Persistence Daemon.

      After the Tesla driver is installed, Persistence-M is off by default. The driver is more stable when Persistence-M is on. To ensure stable service, enable Persistence-M mode using the NVIDIA Persistence Daemon. For more information, see Persistence Daemon.

      Note

      1. Run the following command to start the NVIDIA Persistence Daemon.

        sudo nvidia-persistenced --user username 
# Replace username with your username.

      2. Run the following command to check the status of Persistence-M.

        nvidia-smi

        A response similar to the following example indicates that Persistence-M is on.

        persistence.jpg

    5. (Optional) Enable Persistence-M after a system restart.

      If the system restarts, the on state of Persistence-M is not preserved. You can perform the following steps to re-enable the Persistence-M property.

      The Tesla driver installation package installs the NVIDIA installation scripts, such as sample scripts and installer scripts, to the /usr/share/doc/NVIDIA_GLX-1.0/samples/nvidia-persistenced-init.tar.bz2 path.

      1. Run the following commands to decompress and install the NVIDIA installation scripts.

        cd  /usr/share/doc/NVIDIA_GLX-1.0/samples/
sudo tar xf nvidia-persistenced-init.tar.bz2
cd  nvidia-persistenced-init
sudo sh install.sh

      2. Run the following command to check whether the NVIDIA Persistence Daemon is running.

        sudo systemctl status nvidia-persistenced

        The NVIDIA Persistence Daemon is running if the response is similar to the following example.

        persistence Daemon.jpg

        Note

        You can adapt the NVIDIA Persistence Daemon installation script to ensure that it works correctly on your operating system.

      3. Run the following command to confirm that Persistence-M is on.

        nvidia-smi

      4. (Optional) Run the following commands to stop the NVIDIA Persistence Daemon.

        If you do not need to run the NVIDIA Persistence Daemon, you can stop it.

        sudo systemctl stop nvidia-persistenced
sudo systemctl disable nvidia-persistenced

    6. If your GPU-accelerated instance family is ebmgn8v, ebmgn7, ebmgn7e, install the nvidia-fabricmanager service that corresponds to the driver version.

      Important

      • If the GPU-accelerated instance family is ebmgn8v, ebmgn7, ebmgn7e, the GPU-accelerated instance will not function correctly if the nvidia-fabricmanager service that corresponds to the driver version is not installed.

      • If the GPU-accelerated instance family is not ebmgn8v, ebmgn7, ebmgn7e, skip this step.

      1. Install the nvidia-fabricmanager service.

        You can install the nvidia-fabricmanager service from the source code or from a package. The following examples show the commands for CentOS 7.x and CentOS 8.x. In the following examples, the driver version (driver_version) is 460.91.03. Replace driver_version with the version number of the driver that you downloaded in Step 1: Download the NVIDIA Tesla driver.

        • From source code

          • CentOS 7.x

            driver_version=460.91.03
sudo yum -y install yum-utils
sudo yum-config-manager --add-repo https://developer.download.nvidia.com/compute/cuda/repos/rhel7/x86_64/cuda-rhel7.repo
sudo yum install -y nvidia-fabric-manager-${driver_version}-1

          • CentOS 8.x

            driver_version=460.91.03
driver_version_main=$(echo $driver_version | awk -F '.' '{print $1}')
distribution=rhel8
ARCH=$( /bin/arch )
sudo dnf config-manager --add-repo https://developer.download.nvidia.com/compute/cuda/repos/$distribution/${ARCH}/cuda-$distribution.repo
sudo dnf module enable -y nvidia-driver:${driver_version_main}
sudo dnf install -y nvidia-fabric-manager-0:${driver_version}-1

        • Package-based installation

          • CentOS 7.x

            driver_version=460.91.03
sudo wget https://developer.download.nvidia.com/compute/cuda/repos/rhel7/x86_64/nvidia-fabric-manager-${driver_version}-1.x86_64.rpm
sudo rpm -ivh nvidia-fabric-manager-${driver_version}-1.x86_64.rpm

          • CentOS 8.x

            driver_version=460.91.03
sudo wget https://developer.download.nvidia.com/compute/cuda/repos/rhel8/x86_64/nvidia-fabric-manager-${driver_version}-1.x86_64.rpm
sudo rpm -ivh nvidia-fabric-manager-${driver_version}-1.x86_64.rpm

      2. Run the following commands to start the nvidia-fabricmanager service.

        sudo systemctl enable nvidia-fabricmanager
sudo systemctl start nvidia-fabricmanager

      3. Run the following command to check if the nvidia-fabricmanager service is installed.

        systemctl status nvidia-fabricmanager

        The following output indicates that the nvidia-fabricmanager service is installed.

        Dingtalk_20240910143221.jpg

    Ubuntu and other operating systems

    1. Grant permissions and install the Tesla driver.

      For a driver for Linux 64-bit, use the .run file, such as NVIDIA-Linux-x86_64-xxxx.run. Run the following commands to grant permissions and install the Tesla driver.

      Note

      If you are using a Tesla driver in another format, such as .deb or .rpm, see the NVIDIA CUDA Installation Guide for Linux for installation instructions.

      sudo chmod +x NVIDIA-Linux-x86_64-xxxx.run
      sudo sh NVIDIA-Linux-x86_64-xxxx.run

    2. Run the following command to check whether the Tesla driver is installed.

      nvidia-smi

      A response similar to the following example indicates that the Tesla driver is installed.

      驱动版本.jpg

    3. (Optional) Enable Persistence-M mode using the NVIDIA Persistence Daemon.

      After the Tesla driver is installed, Persistence-M is off by default. The driver is more stable when Persistence-M is on. To ensure stable service, enable Persistence-M mode using the NVIDIA Persistence Daemon. For more information, see Persistence Daemon.

      Note

      1. Run the following command to start the NVIDIA Persistence Daemon.

        sudo nvidia-persistenced --user username 
# Replace username with your username.

      2. Run the following command to check the status of Persistence-M.

        nvidia-smi

        A response similar to the following example indicates that Persistence-M is on.

        persistence.jpg

    4. (Optional) Enable Persistence-M after a system restart.

      If the system restarts, the on state of Persistence-M is not preserved. You can perform the following steps to re-enable the Persistence-M property.

      The Tesla driver installation package installs the NVIDIA installation scripts, such as sample scripts and installer scripts, to the /usr/share/doc/NVIDIA_GLX-1.0/samples/nvidia-persistenced-init.tar.bz2 path.

      1. Run the following commands to decompress and install the NVIDIA installation scripts.

        cd  /usr/share/doc/NVIDIA_GLX-1.0/samples/
sudo tar xf nvidia-persistenced-init.tar.bz2
cd  nvidia-persistenced-init
sudo sh install.sh

      2. Run the following command to check whether the NVIDIA Persistence Daemon is running.

        sudo systemctl status nvidia-persistenced

        The NVIDIA Persistence Daemon is running if the response is similar to the following example.

        persistence Daemon.jpg

        Note

        You can adapt the NVIDIA Persistence Daemon installation script to ensure that it works correctly on your operating system.

      3. Run the following command to confirm that Persistence-M is on.

        nvidia-smi

      4. (Optional) Run the following commands to stop the NVIDIA Persistence Daemon.

        If you do not need to run the NVIDIA Persistence Daemon, you can stop it.

        sudo systemctl stop nvidia-persistenced
sudo systemctl disable nvidia-persistenced

    5. If your GPU-accelerated instance family is ebmgn8v, ebmgn7, ebmgn7e, install the nvidia-fabricmanager service that corresponds to the driver version.

      Important

      • If the GPU-accelerated instance family is ebmgn8v, ebmgn7, ebmgn7e, the GPU-accelerated instance will not function correctly if the nvidia-fabricmanager service that corresponds to the driver version is not installed.

      • If the GPU-accelerated instance family is not ebmgn8v, ebmgn7, ebmgn7e, skip this step.

      1. Install the nvidia-fabricmanager service.

        You can install the nvidia-fabricmanager service from the source code or from a package. The following examples show the commands for Ubuntu 16.04, Ubuntu 18.04, Ubuntu 20.04, Ubuntu 22.04, and Ubuntu 24.04. Replace driver_version with the version number of the driver that you downloaded in Step 1: Download the NVIDIA Tesla driver.

        Important

        • To install the nvidia-fabricmanager service on Ubuntu 22.04, the Tesla driver version must be 515.48.07 or later. The following sample code for Ubuntu 22.04 uses driver version 535.154.05 as an example.

        • To install the nvidia-fabricmanager service on Ubuntu 24.04, the Tesla driver version must be 550.90.07 or later. The following sample code for Ubuntu 24.04 uses driver version 570.133.20 as an example.

        • From source code

          Ubuntu 16.04, Ubuntu 18.04, or Ubuntu 20.04

          driver_version=460.91.03
driver_version_main=$(echo $driver_version | awk -F '.' '{print $1}')
distribution=$(. /etc/os-release;echo $ID$VERSION_ID | sed -e 's/\.//g')
sudo wget https://developer.download.nvidia.com/compute/cuda/repos/$distribution/x86_64/cuda-$distribution.pin
sudo mv cuda-$distribution.pin /etc/apt/preferences.d/cuda-repository-pin-600
sudo wget https://developer.download.nvidia.com/compute/cuda/repos/$distribution/x86_64/3bf863cc.pub
sudo apt-key add 3bf863cc.pub
sudo rm 3bf863cc.pub
sudo echo "deb https://developer.download.nvidia.com/compute/cuda/repos/$distribution/x86_64 /" | tee /etc/apt/sources.list.d/cuda.list
sudo apt-get update
sudo apt-get -y install nvidia-fabricmanager-${driver_version_main}=${driver_version}-*

          Ubuntu 22.04

          driver_version=535.154.05
driver_version_main=$(echo $driver_version | awk -F '.' '{print $1}')
distribution=$(. /etc/os-release;echo $ID$VERSION_ID | sed -e 's/\.//g')
sudo wget https://developer.download.nvidia.com/compute/cuda/repos/$distribution/x86_64/cuda-$distribution.pin
sudo mv cuda-$distribution.pin /etc/apt/preferences.d/cuda-repository-pin-600
sudo wget https://developer.download.nvidia.com/compute/cuda/repos/$distribution/x86_64/3bf863cc.pub
sudo apt-key add 3bf863cc.pub
sudo rm 3bf863cc.pub
sudo echo "deb https://developer.download.nvidia.com/compute/cuda/repos/$distribution/x86_64 /" | tee /etc/apt/sources.list.d/cuda.list
sudo apt-get update
sudo apt-get -y install nvidia-fabricmanager-${driver_version_main}=${driver_version}-*

          Ubuntu 24.04

          driver_version=570.133.20
driver_version_main=$(echo $driver_version | awk -F '.' '{print $1}')
distribution=$(. /etc/os-release;echo $ID$VERSION_ID | sed -e 's/\.//g')
sudo wget https://developer.download.nvidia.com/compute/cuda/repos/$distribution/x86_64/cuda-$distribution.pin
sudo mv cuda-$distribution.pin /etc/apt/preferences.d/cuda-repository-pin-600
sudo wget https://developer.download.nvidia.com/compute/cuda/repos/$distribution/x86_64/3bf863cc.pub
sudo apt-key add 3bf863cc.pub
sudo rm 3bf863cc.pub
sudo echo "deb https://developer.download.nvidia.com/compute/cuda/repos/$distribution/x86_64 /" | tee /etc/apt/sources.list.d/cuda.list
sudo apt-get update
sudo apt-get -y install nvidia-fabricmanager-${driver_version_main}=${driver_version}-*

        • From a package

          • Ubuntu 16.04

            driver_version=460.91.03
driver_version_main=$(echo $driver_version | awk -F '.' '{print $1}')
sudo wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu1604/x86_64/nvidia-fabricmanager-${driver_version_main}_${driver_version}-1_amd64.deb
sudo dpkg -i nvidia-fabricmanager-${driver_version_main}_${driver_version}-1_amd64.deb

          • Ubuntu 18.04

            driver_version=460.91.03
driver_version_main=$(echo $driver_version | awk -F '.' '{print $1}')
sudo wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu1804/x86_64/nvidia-fabricmanager-${driver_version_main}_${driver_version}-1_amd64.deb
sudo dpkg -i nvidia-fabricmanager-${driver_version_main}_${driver_version}-1_amd64.deb

          • Ubuntu 20.04

            driver_version=460.91.03
driver_version_main=$(echo $driver_version | awk -F '.' '{print $1}')
sudo wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2004/x86_64/nvidia-fabricmanager-${driver_version_main}_${driver_version}-1_amd64.deb
sudo dpkg -i nvidia-fabricmanager-${driver_version_main}_${driver_version}-1_amd64.deb

          • Ubuntu 22.04

            driver_version=535.154.05 
driver_version_main=$(echo $driver_version | awk -F '.' '{print $1}')
sudo wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/nvidia-fabricmanager-${driver_version_main}_${driver_version}-1_amd64.deb
sudo dpkg -i nvidia-fabricmanager-${driver_version_main}_${driver_version}-1_amd64.deb

          • Ubuntu 24.04

            driver_version=570.133.20 
driver_version_main=$(echo $driver_version | awk -F '.' '{print $1}')
sudo wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2404/x86_64/nvidia-fabricmanager-${driver_version_main}_${driver_version}-1_amd64.deb
sudo dpkg -i nvidia-fabricmanager-${driver_version_main}_${driver_version}-1_amd64.deb

      2. Run the following commands to start the nvidia-fabricmanager service.

        sudo systemctl enable nvidia-fabricmanager
sudo systemctl start nvidia-fabricmanager

      3. Run the following command to check whether the nvidia-fabricmanager service is installed.

        systemctl status nvidia-fabricmanager

        The following response indicates that the nvidia-fabricmanager service is installed.

        image.png

        Note

        The nvidia-fabricmanager package version must match the Tesla driver version to ensure that the GPU works correctly. On Ubuntu, if you install the nvidia-fabricmanager service from a package, the apt-daily service might automatically update the package. This can cause a version mismatch between the nvidia-fabricmanager package and the Tesla driver. As a result, the nvidia-fabricmanager service fails to start and the GPU becomes unusable. For more information about how to resolve this issue, see The GPU is unusable because the nvidia-fabricmanager version is inconsistent with the Tesla driver version.

    References