Conda ships Intel MKL for linear algebra, too, meaning numpy.dot (matrix multiply), scipy, fft, solvers, … run 2-10x faster. Well, used to, I think the open sou

Conda ships Intel MKL for linear algebra, too, meaning numpy.dot (matrix multiply), scipy, fft, solvers, … run 2-10x faster. Well, used to, I think the open source alternatives have improved since.