Computer_Vision/Chapter02/Numpy_Vs_Torch_object_computation_speed_comparison.ipynb

import torch
x = torch.rand(1, 6400)
y = torch.rand(6400, 5000)

device = 'cuda' if torch.cuda.is_available() else 'cpu'
assert device == 'cuda', "This exercise assumes the notebook is on a GPU machine"

x, y = x.to(device), y.to(device)

%timeit z=(x@y)

The slowest run took 22.35 times longer than the fastest. This could mean that an intermediate result is being cached.
10000 loops, best of 3: 974 µs per loop

x, y = x.cpu(), y.cpu()
%timeit z=(x@y)

100 loops, best of 3: 9.4 ms per loop

import numpy as np
x = np.random.random((1, 6400))
y = np.random.random((6400, 5000))
%timeit z = np.matmul(x,y)

10 loops, best of 3: 19.9 ms per loop