failed call to cuInit: CUDA_ERROR_NO_DEVICE: no CUDA-capable device is detected

Question

Whenever I import tensorflow in my code, I get the following warning

2021-07-03 20:43:38.432690: E tensorflow/stream_executor/cuda/cuda_driver.cc:313] failed call to cuInit: CUDA_ERROR_NO_DEVICE: no CUDA-capable device is detected

To check gpu availability I ran

tf.config.list_physical_devices('GPU')

For which it returned []

I also ran

tf.test.is_gpu_available()

Which returned some noticeable things

2021-07-03 20:49:02.552671: I tensorflow/core/platform/cpu_feature_guard.cc:143] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 FMA

2021-07-03 20:49:02.584748: I tensorflow/core/platform/profile_utils/cpu_utils.cc:102] CPU Frequency: 2394575000 Hz

2021-07-03 20:49:02.585433: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x7f32e8000b60 initialized for platform Host (this does not guarantee that XLA will be used). Devices:

2021-07-03 20:49:02.585504: I tensorflow/compiler/xla/service/service.cc:176]   StreamExecutor device (0): Host, Default Version

2021-07-03 20:49:02.590168: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcuda.so.1

2021-07-03 20:49:02.737950: E tensorflow/stream_executor/cuda/cuda_driver.cc:313] failed call to cuInit: CUDA_ERROR_NO_DEVICE: no CUDA-capable device is detected

2021-07-03 20:49:02.737996: I tensorflow/stream_executor/cuda/cuda_diagnostics.cc:156] kernel driver does not appear to be running on this host (user_name-HP-Pavilion-15-Notebook-PC): /proc/driver/nvidia/version does not exist
False

Following are my setup specifications

Python 3.8.5
TensorFlow 2.2.0
nvcc 10.1
os Ubuntu 20.04 LTS
GPU GM108M [GeForce 830M]

What should I do to resolve this? I want tensorflow lib. to utilise the local GPU while running the code.

EDIT: libcudart,libcuda are installed.

Answer 1

As @Dr. Snoopy mentioned, Nvidia graphics drivers were missing. I installed them using link, https://www.nvidia.com/Download/index.aspx?lang=en-us

I also installed, CuDNN from https://developer.nvidia.com/rdp/cudnn-archive

Now, when I type

tf.config.list_physical_devices('GPU')

I get

[PhysicalDevice(name='/physical_device:GPU:0',device_type='GPU')]