I really enjoy using MKL libraries since they are much faster on our 36 core Intel(R) Xeon(R) Gold 6140 CPU @ 2.30GHz mashine. However our scientific research requires diagonalization of very large matrices which consume a lot of memory and MKL scalapack uses much more memory compared to NETLIB version. I am attaching graph where i show memory consumption of our code using intel and netlib libraries. Diagonalisation (pdsyevr) is performed between vertical lines. One can notice sudden leap of used memory in the middle of diagonalisation procedure in MKL version. Is it possible somehow to get rid of this behavior since it drains all RAM.