Python NumPy - new matrix multiplication operator @ vs numpy.matmul - are they always the same?

Question

I just ran across the new (as of Python 3.5 apparently) matrix multiplication operator @ used with 2 NumPy matrices. Previously, when performing matrix multipl