Benchmark GPU vs CPU with TensorFlow

Benchmark GPU vs CPU & multi-host vs single host

This notebook can be used to benchmark performance using CPU, a single GPU or many GPUs.

Tested with TensorFlow 2.4.0

Machine Learning on Hopsworks


The hops python module

hops is a helper library for Hops that facilitates development by hiding the complexity of running applications and iteracting with services.

Have a feature request or encountered an issue? Please let us know on github.

Using the experiment module

To be able to run your Machine Learning code in Hopsworks, the code for the whole program needs to be provided and put inside a wrapper function. Everything, from importing libraries to reading data and defining the model and running the program needs to be put inside a wrapper function.

The experiment module provides an api to Python programs such as TensorFlow, Keras and PyTorch on a Hopsworks on any number of machines and GPUs.

An Experiment could be a single Python program, which we refer to as an Experiment.

Grid search or genetic hyperparameter optimization such as differential evolution which runs several Experiments in parallel, which we refer to as Parallel Experiment.

ParameterServerStrategy, CollectiveAllReduceStrategy and MultiworkerMirroredStrategy making multi-machine/multi-gpu training as simple as invoking a function for orchestration. This mode is referred to as Distributed Training.

Using the tensorboard module

The tensorboard module allow us to get the log directory for summaries and checkpoints to be written to the TensorBoard we will see in a bit. The only function that we currently need to call is tensorboard.logdir(), which returns the path to the TensorBoard log directory. Furthermore, the content of this directory will be put in as a Dataset in your project’s Experiments folder.

The directory could in practice be used to store other data that should be accessible after the experiment is finished.

# Use this module to get the TensorBoard logdir
from hops import tensorboard
tensorboard_logdir = tensorboard.logdir()

Using the hdfs module

The hdfs module provides a method to get the path in HopsFS where your data is stored, namely by calling hdfs.project_path(). The path resolves to the root path for your project, which is the view that you see when you click Data Sets in HopsWorks. To point where your actual data resides in the project you to append the full path from there to your Dataset. For example if you create a mnist folder in your Resources Dataset, the path to the mnist data would be hdfs.project_path() + 'Resources/mnist'

# Use this module to get the path to your project in HopsFS, then append the path to your Dataset in your project
from hops import hdfs
project_path = hdfs.project_path()
# Downloading the mnist dataset to the current working directory
from hops import hdfs
mnist_hdfs_path = hdfs.project_path() + "Resources/mnist"
local_mnist_path = hdfs.copy_to_local(mnist_hdfs_path)


See the following links to learn more about running experiments in Hopsworks

Managing experiments

Experiments service provides a unified view of all the experiments run using the experiment module.
As demonstrated in the gif it provides general information about the experiment and the resulting metric. Experiments can be visualized meanwhile or after training in a TensorBoard.


def wrapper():
    import tensorflow as tf
    # Wrapper for keras_applications, you can import any model you want to try (like ResNet50)
    from tensorflow.keras.applications import ResNet50

    import numpy as np
    from hops import tensorboard
    # Utility module for getting number of GPUs accessible by the container (Spark Executor)
    from hops import devices

    batch_size = 8 # Number of samples to process on each GPU every iteration
    # Image dimensions
    height = 224
    width = 224
    channels = 3
    num_classes = 1000
    num_iterations = 5000 # Number of iterations, increase to run longer

    log_dir = tensorboard.logdir()
    # Read synthetic data (can be replaced with real data)
    def input_fn():
      data = np.random.random((batch_size, height, width, channels)).astype(np.float32)
      labels = np.random.random((batch_size, num_classes))
      dataset =, labels))
      dataset = dataset.repeat(num_iterations)
      dataset = dataset.batch(batch_size)
      return dataset    

    # Define distribution strategy
    strategy = tf.distribute.MirroredStrategy()

    with strategy.scope():
        model = ResNet50(weights=None, input_shape=(height, width, channels), classes=num_classes)

        optimizer = tf.keras.optimizers.RMSprop(0.2)
        model.compile(loss='categorical_crossentropy', optimizer=optimizer)

    callbacks = [
from hops import experiment
experiment.launch(wrapper, local_logdir=True)
Finished Experiment 

('hdfs://rpc.namenode.service.consul:8020/Projects/demo/Experiments/application_1594231828166_0163_3', {'metric': None, 'log': 'Experiments/application_1594231828166_0163_3/output.log'})