Native numpy parallelization - multiplication and sum/average

Question

Is there a way to enable automatic parallelization for basic numpy operations, like element-wise multiplication of arrays and basic numpy functions like np.sum and np.average?

I know that it is possible for blas/lapack functions, as discussed for scipy.linalg.solve in this thread:

Is it possible to know which SciPy / NumPy functions run on multiple cores?

And I managed to run this code natively in parallel via MKL:

import numpy

def test():
    n = 5000
    data = numpy.random.random((n, n))
    result = numpy.linalg.inv(data)

test();

But I would need to run something like this in parallel:

   N = 1024    
   A = np.zeros((N,N,N),dtype='float32')    
   B = np.zeros((N,N,N),dtype='float32')
   C = np.zeros((N,N,N),dtype='float32')

   A[:,:,:] = 1
   B[:,:,:] = 2

   # this is the part I want parallel
   C[:,:,:] = A[:,:,:]*B[:,:,:]

   # also this:
   avgC = np.average(C)

Otherwise, what would be the simplest way to paralelize these target operations?

Answer 1

There are several automatic parallelization tools for Python, but I'm not sure if they are compatible with Numpy. It might be possible to parallelize these operations by combining a Python-to-C++ translator with an automatic parallelization tool for C++.