Are numpy's basic operations vectorized, i.e. do they use SIMD operations?

Question

I am doing some performance analysis, and i wonder, whether numpy vectorizes its standard array operations, when the datatype is known (double).

Answer 1

Yes, they are.

/*
 * This file is for the definitions of simd vectorized operations.
 *
 * Currently contains sse2 functions that are built on amd64, x32 or
 * non-generic builds (CFLAGS=-march=...)
 * In future it may contain other instruction sets like AVX or NEON     detected
 * at runtime in which case it needs to be included indirectly via a file
 * compiled with special options (or use gcc target attributes) so the binary
 * stays portable.
 */

Link: Numpy simd.inc.src on github.

Answer 2

Take a look at basic example

import numpy as np

x = np.array([1, 2, 3], np.int32)
print (type(x))
y = np.array([6, 7, 8], np.int32)
print (type(y))

Now we are adding up these two arrays

z=x+y
print (z)
print (type(z))

As a result we have

<class 'numpy.ndarray'>
<class 'numpy.ndarray'>
[ 7  9 11]
<class 'numpy.ndarray'>

Vectorised,yes they are.But the term vector has different meaning in mathematics and physics,and we are using arrays as mathematical abstraction.