pyVideoTrans

English

简体中文

English

简体中文

Appearance

If you don't need CUDA acceleration, you don't need to install and configure CUDA, cuDNN, etc. ​

CUDA is a toolkit developed by NVIDIA that allows your graphics card to handle more tasks. In this software (pyVideoTrans), if you want to enable CUDA acceleration, you must first install the CUDA environment. This article will guide you step-by-step through installing CUDA 12.6 and the corresponding cuDNN 9.8 on Windows 10.

Why choose CUDA 12.6 instead of 12.8? Because 12.8 is the latest version and may not be well adapted by some software yet, which can easily lead to compatibility issues. On the other hand, 12.6 is a relatively stable and up-to-date version. Of course, if you want 12.4 or 12.8, the installation method is similar, and you can refer to this tutorial and adjust the version number.

Step 1: Check if your graphics card is NVIDIA ​

  1. Open Device Manager

    • Right-click on the "Start" button (the Windows icon) in the lower-left corner of your desktop.
    • In the pop-up menu, select "Windows Tools - Computer Management" to open a window.

  2. Find graphics card information

    • In the Computer Management window, click Device Manager on the left side, find "Display adapters" on the right side, and click the small arrow in front of it to expand.

    • See if there is "NVIDIA" in it, such as "NVIDIA GeForce GTX 1660".

    • If so, congratulations, your graphics card is NVIDIA, and you can continue. If not (for example, it shows "Intel" or "AMD"), then CUDA cannot be used, and the tutorial ends here.

Step 2: Check and update the graphics card driver version ​

  1. Open NVIDIA Management Software

    • If you have installed the graphics card driver, there may be a green "NVIDIA" icon in the taskbar (next to the time) in the lower right corner of the desktop. Right-click it and select "NVIDIA GeForce Experience" to open it.

    • If not, skip to step 3 to download it manually.

  2. Check and update the driver version

    • After opening, click "Drivers" in the upper left corner.
    • It will display the current driver version (such as "546.33"). If there is a new version, there will be a "Download" button.
    • Click "Download", and then follow the prompts to install it. Restart your computer after installation.

  3. No software? Manually download the driver

    • Open your browser and enter the website: https://www.nvidia.com/en-us/geforce/drivers/ to download the driver upgrade software, and then install it.
    • After installation, open it and follow the instructions to check for updated drivers.

Step 3: Check if CUDA is already installed and determine the maximum supported version ​

  1. Open the command line terminal (CMD)
    • Hold down the Windows key + R on your keyboard to bring up the "Run" window.
    • Enter cmd and press Enter to open a black command line window.

  1. Check CUDA version
    • Enter in the command line: nvcc -V

Press Enter. If it displays something like Cuda compilation tools, release 12.6, V12.6.xxx, it means that CUDA has been installed and the version is 12.6. If it displays "is not an internal or external command", it means that it has not been installed, so continue with the installation.

If the version is 11.x, please uninstall it and reinstall 12.x.

  1. View the maximum supported CUDA version
    • Enter in the command line: nvidia-smi

Press Enter, a table will be displayed, and there is "CUDA Version" in the upper right corner (such as 12.6 or higher). This means the highest version supported by your driver. As long as it is 12.6 or higher, you can install the CUDA 12.6 that we want. If it shows 11.x, it means the graphics card driver is too old, please update the graphics card driver first.

Step 4: Download and install CUDA 12.6 ​

  1. Download CUDA 12.6

  1. Install CUDA (custom mode)

    • Double-click the downloaded file, a installation window will pop up, agree to the agreement and click Next.

    • Select "Custom (Advanced)" installation and click "Next".

    • In the options list, only check "CUDA" in the first line, and uncheck all other options to avoid conflicts.

  • Then click the + sign in front of CUDA to uncheck Visual Studio Integration, otherwise the installation may fail.

  • Click "Next" and follow the prompts to complete the installation. The default path is C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.6.

  1. What if there is an error?

    • If the installation is interrupted and prompts "Not installed" or "Failed", it may be that the option is not selected correctly. Re-run the installation program, select custom installation, make sure to only select "CUDA" in the first line, uncheck all others, and open CUDA to uncheck Visual Studio Integration.
    • If the error still occurs, skip to the next step to install Visual Studio.

Step 5: Install Visual Studio (if CUDA installation fails) ​

If you unchecked Visual Studio Integration in the previous step, there should be no errors. If an error occurs, please install Visual Studio according to these steps.

  1. Download Visual Studio Community Edition

  2. Install C++ development environment

    • Double-click the installer. After the window pops up, select "Desktop development with C++".
    • Click "Install", which may take tens of minutes. Restart your computer after installation.

  1. Reinstall CUDA 12.6
    • Go back to step 4 and re-run the CUDA installer. Select "Runtime" installation in custom mode. This time it should be successful.

Step 6: Verify CUDA installation and configure environment variables ​

  1. Check if the installation is successful

    • Open CMD (Windows key + R, enter cmd).
    • Enter: nvcc -V If it displays Cuda compilation tools, release 12.6, V12.6.xxx, it means the installation is successful. If it prompts "is not an internal or external command", you need to add environment variables.

  2. Add CUDA to environment variables (automatically set by default, if not, please add manually)

    • Right-click "This PC", select "Properties" > "Advanced system settings" > "Environment Variables".
    • In the "Path" of "System variables", click "New", and enter:
      C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.6\bin
C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.6\lib
C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.6\include
C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.6\libnvvp
    • Click "OK" to save and close all windows.
    • Reopen CMD and enter nvcc -V. You should be able to see the version information.

Step 7: Install cuDNN 9.8 ​

  1. Download cuDNN 9.8

  1. Install cuDNN
    • Double-click the downloaded .exe file and follow the prompts to install it.

If cuDNN cannot be used in the program after installation, you can open the installation directory, the default is C:\Program Files\NVIDIA\CUDNN\v9.8, copy the lib bin include 3 folders, copy them to C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.6 to overwrite.

Finally: You're done! ​

Now CUDA 12.6 and cuDNN 9.8 are installed! You can start running programs that require GPU support. If there are software instructions (such as running python script.py), enter the project folder in CMD and enter the corresponding command to try it out.