How to tell if TensorFlow is using GPU acceleration from inside Python shell

If you are working with TensorFlow, you may want to see if the library is using GPU acceleration for your machine learning tasks. There are a few ways to check this from inside the Python shell.

Check for GPU availability

First, you can check if a GPU is available for TensorFlow to use. Here’s a simple way to do that:


import tensorflow as tf
print(tf.config.list_physical_devices('GPU'))


If a GPU is available, you should see an output indicating its presence. If not, then TensorFlow is not currently using GPU acceleration.

Check for GPU usage

After confirming the presence of a GPU, you can also check if it is being used by TensorFlow for computations. To do this, you can use the following code:


import tensorflow as tf
print(tf.test.is_built_with_cuda())
print(tf.test.is_gpu_available(cuda_only=False, min_cuda_compute_capability=None))


The first line will print out whether TensorFlow is built with CUDA support, while the second line will indicate if the GPU is currently available for computations. If both return True, then TensorFlow is successfully using GPU acceleration.

Additional information

If you want to delve deeper into the details of GPU usage in TensorFlow, you can also use the following code to print out a summary of the GPU utilization:


import tensorflow as tf
print(tf.config.experimental.get_memory_growth('GPU:0'))
print(tf.config.experimental.get_virtual_device_configuration())


This will provide you with more specific information about how TensorFlow is utilizing the GPU for your tasks.

By using these methods, you can easily determine if TensorFlow is using GPU acceleration from inside the Python shell, giving you insight into the computational power being leveraged for your machine learning work.