CUDA, cuDNN for Tensorflow on Ubuntu


I’m trying to make a Convolutional Neural Network for image classification using Tensorflow 2.0 and cuDNN 7.6.4.

Build information:
UBUNTU 18.04
CUDA 10.0
NVIDIA Driver Version 430.40 (from command prompt line nvidia-smi)
Tensorflow 2.0

I have been having issues installing the cuDNN version for CUDA 10.0 (cuDNN 7.6.4 for CUDA 10.0). I have followed the documentation trying the .deb files and the .tgz file, but when doing the verification instructions in the documentation, I get an error on the “make clean && make” command: “skipping incompatible /usr/local/cuda/lib64/ when searching for -lcudnn”.

Are there any fixes for this problem or are there any known issues with the combination of hardware/software in the list above?

Could you please let us know if you are still facing this issue?