#7061 CUDA GPU acceleration unavailable. Verify NVIDIA GPU & CUDA installation (Help -> CUDA Help).

This error means the software could not detect an NVIDIA GPU or a compatible CUDA/cuDNN environment, so GPU acceleration cannot be used. This is not a software bug — it's a system-level dependency issue.

Steps to resolve:

1. Check if you have an NVIDIA GPU

   • GPU acceleration only works with NVIDIA GPUs. AMD/Intel GPUs are not supported.
   • Press Win + R, type dxdiag, and check the "Display" tab. If you see a non-NVIDIA GPU or only an integrated graphics card, CUDA acceleration is impossible.

2. Verify CUDA version

   • Open Command Prompt (CMD) and run:

     nvcc -V

   • The output must show version 12.8 or higher. If nvcc is not found, CUDA Toolkit is not installed.

3. Verify cuDNN version

   • cuDNN must be version 9.11 or higher. It is not included in nvcc -V — check the bin and lib directories under your CUDA installation path.

4. Check NVIDIA driver

   • Run:

     nvidia-smi

   • This shows your driver version and the highest CUDA version it supports. Ensure the driver is recent enough to support CUDA 12.8+.

5. Configure environment variables

   • If CUDA/cuDNN is installed but you still see this error, verify the PATH environment variable includes:

     • C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.8\bin
     • C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.8\lib
   • Add them if missing, then restart sp.exe.

6. If you do not have an NVIDIA GPU

   • Simply uncheck the "CUDA加速" option in the main interface. The software will fall back to CPU processing, which is slower but fully functional.

Additional notes:

• The pre-packaged version does not come with CUDA/cuDNN — you must install them separately following the guide.
• For detailed installation steps, see the official CUDA/cuDNN setup document.

Please refer to the relevant documentation:

• Configuring CUDA and cuDNN on Windows
• CUDA Help