How to use Tensorflow with AMD Vega 10 GPU

Mike Evans 0 Reputation points
2023-10-20T17:40:13.8966667+00:00

Hello,

I have been using Azure virtual machines to train deep learning models in Tensorflow. Recently, I created an NVas v4 machine with the Data Science VM Ubuntu image and am unable to use the GPU. My understanding is that the NVas v4 GPU is an AMD Vega 10 (as indicated by the output of lshw -class display). Therefore, because the GPU is not made by NVIDIA the usual drivers and libraries that Tensorflow would use to recognize the GPU (CUDA, cuDNN) are irrelevant. Are there additional packages/libraries needed to use Tensorflow with an AMD GPU? It seems odd that Microsoft would offer VMs with GPUs that do not have the necessary software to be utilized.

Mike

Azure Virtual Machines
Azure Virtual Machines
An Azure service that is used to provision Windows and Linux virtual machines.
7,975 questions
Azure Data Science Virtual Machines
Azure Data Science Virtual Machines
Azure Virtual Machine images that are pre-installed, configured, and tested with several commonly used tools for data analytics, machine learning, and artificial intelligence training.
69 questions
0 comments No comments
{count} votes

1 answer

Sort by: Most helpful
  1. Vahid Ghafarpour 21,725 Reputation points
    2023-10-20T17:44:56.42+00:00

    To use Tensorflow on AMD GPUs, you can use ROCm.

    https://github.com/RadeonOpenCompute/ROCm


Your answer

Answers can be marked as Accepted Answers by the question author, which helps users to know the answer solved the author's problem.