TensorFlow 2 - CPU vs GPU times calculation on matrix multiplications

CPUvsGPU_tf2.py

from __future__ import print_function
import matplotlib
import matplotlib.pyplot as plt
import tensorflow as tf
import time
import os
tf.compat.v1.disable_eager_execution()
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3'

def get_times(maximum_time):

    device_times = {
        "/gpu:0":[],
        "/cpu:0":[]
    }
    matrix_sizes = range(500,50000,100)

    for size in matrix_sizes:
        print("####### Matrix size: " + str(size) + " #######")
        for device_name in device_times.keys():

            print("####### Calculating on the " + device_name + " #######")

            shape = (size,size)
            data_type = tf.float16
            with tf.device(device_name):
                r1 = tf.random.uniform(shape=shape, minval=0, maxval=1, dtype=data_type)
                r2 = tf.random.uniform(shape=shape, minval=0, maxval=1, dtype=data_type)
                dot_operation = tf.matmul(r2, r1)

            with tf.compat.v1.Session(config=tf.compat.v1.ConfigProto(log_device_placement=False)) as session:
                    start_time = time.time()
                    result = session.run(dot_operation)
                    time_taken = time.time() - start_time
                    #print(result)
                    device_times[device_name].append(time_taken)
            print("Time taken:", time_taken)

            if time_taken > maximum_time:
                return device_times, matrix_sizes

device_times, matrix_sizes = get_times(10) # Change the number to specify maximum cut-off compute time (in seconds) after which the comparison script is terminated and results displayed.
print(device_times)
gpu_times = device_times["/gpu:0"]
cpu_times = device_times["/cpu:0"]

plt.plot(matrix_sizes[:len(gpu_times)], gpu_times, 'o-')
plt.plot(matrix_sizes[:len(cpu_times)], cpu_times, 'o-')
plt.ylabel('Time')
plt.xlabel('Matrix size')
plt.show()
plt.plot(matrix_sizes[:len(cpu_times)], [a/b for a,b in zip(cpu_times,gpu_times)], 'o-')
plt.ylabel('CPU Time / GPU Time')
plt.xlabel('Matrix size')
plt.show()