0

I am experienced in machine learning and would like to attempt working with AlphaFold. However, I am encountering some difficulties. To begin, I have created a virtual environment and proceeded to install CuDNN (version 8.1.0.77), CuDNN Toolkit (version 11.2.2), TensorFlow (version 2.10.0), and TensorFlow-GPU (version 2.10.0). After completing these installations, I verified the versions using nvidia-smi and nvcc --version.

However, I have noticed disparities in the version numbers. In particular, the CUDA version displayed by nvidia-smi is 11.4, while the version indicated by nvcc is 10.1. Additionally, the version of CuDNN Toolkit appears as 11.2. I am uncertain about the relationships between these versions and whether there is a need to rectify this situation. I am seeking guidance on understanding and addressing this issue. Your assistance in clarifying this matter would be greatly appreciated.

$ nvidia-smi
    +-----------------------------------------------------------------------------+
| NVIDIA-SMI 470.199.02   Driver Version: 470.199.02   CUDA Version: 11.4     |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|                               |                      |               MIG M. |
|===============================+======================+======================|
|   0  NVIDIA GeForce ...  Off  | 00000000:04:00.0  On |                  N/A |
|  0%   35C    P8    22W / 420W |    324MiB / 24265MiB |      1%      Default |
|                               |                      |                  N/A |
+-------------------------------+----------------------+----------------------+
                                                                               
+-----------------------------------------------------------------------------+
| Processes:                                                                  |
|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |
|        ID   ID                                                   Usage      |
|=============================================================================|
|    0   N/A  N/A      1060      G   /usr/lib/xorg/Xorg                 66MiB |
|    0   N/A  N/A      1505      G   /usr/bin/gnome-shell               85MiB |
|    0   N/A  N/A      1917      G   /usr/lib/firefox/firefox          170MiB |
+-----------------------------------------------------------------------------+

$ nvcc --version
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2019 NVIDIA Corporation
Built on Sun_Jul_28_19:07:16_PDT_2019
Cuda compilation tools, release 10.1, V10.1.243

My GPU is RTX 3090.

  • Does this answer your question? [Different CUDA versions shown by nvcc and NVIDIA-smi](https://stackoverflow.com/questions/53422407/different-cuda-versions-shown-by-nvcc-and-nvidia-smi) – mhenning Aug 25 '23 at 22:18

0 Answers0