!BLAS/LAPACK

!Ordinarly Least Squares

!Eigenvalue Computation

|-

Again, it should be noted that if MKL were to be running on an AMD platform without bypassing the hardware check, it would perform slower than OpenBLAS. However, while the potential decrease in performance should be acknowledged, given the ability to circumvent it the maximum possible performance should be taken into consideration for the comparison instead.

== Advantages ==