Dear all,
what is the best way of using MKL (Intel's Math Kernel Library) in Fenics?
I'd like to optimize (and possibly parallelize) matrix/vector operations, eigensolvers, and linear solvers.
Thanks!