Mastering Numerical Computing with NumPy

Intel Math Kernel Library

Intel MKL optimizes BLAS for Intel processors. There are improved routines and functions such as Level 1, 2 and 3 BLAS, LAPACK routines, solvers, FFT functions, other math and statistical functions. These improved routines and functions benefit from improvements like shared memory multiprocessing and they are used to accelerate scientific python libraries such as NumPy and SciPy in distributions such as Anaconda Distribution. If you look at its release notes (https://software.intel.com/en-us/articles/intel-math-kernel-library-release-notes-and-new-features), you can see several important improvements were made with every release, such as the improved performance of LAPACK functions.