Dear NVIDIA support team,
I am writing to seek help with my NVIDIA GeForce RTX 3050 GPU, which I’m trying to use for deep learning with PyTorch on a Windows 11 Pro system. The problem is that PyTorch cannot detect the GPU, even though I have installed the latest GPU driver (version 528.49) and PyTorch with CUDA support (version 1.13.1 with CUDA 11.7).
To troubleshoot this issue, I have tried various steps such as reinstalling the GPU driver and PyTorch, checking the compatibility of the GPU and CUDA versions, and verifying the environment variables for CUDA and nvcc. However, I could not find the path to nvcc using the “where nvcc” command, and the “nvcc” command returned an error message “No input files specified”.
I have also checked the CUDA compute capability of my GPU, which is 8.6 according to the NVIDIA website. As far as I know, this should be compatible with CUDA 11.7 and PyTorch 1.13.1. However, PyTorch still cannot detect the GPU, and “torch.cuda.is_available()” returns False.
In addition, I have encountered a problem with the Anaconda prompt, where the “where nvcc” command could find the path to nvcc, but the PyTorch still cannot detect the GPU.
Therefore, I would appreciate it if you could advise me on how to resolve this issue and make PyTorch recognize my GPU. If possible, could you provide me with a step-by-step guide to configure my system for deep learning with PyTorch and CUDA, including the recommended versions of CUDA, PyTorch, and other software? Any help or suggestion would be greatly appreciated.
Thank you for your attention and support.
Sincerely,