Intel® Math Kernel Library 2019 Developer Reference - C
The Extended Eigensolver functionality in Intel Math Kernel Library (Intel MKL) supports two types of partial eigenvalue problem:
Intel MKL only supports the shared memory programming (SMP) version of the eigenvalue solver.
Routines to find the largest/smallest eigenvalues support only symmetric real sparse (CSR, BSR) matrices.
Routines to find the eigenvalues within an interval support only real and complex Hermitian matrices.
The FEAST Algorithm gives a brief description of the algorithm underlying the Extended Eigensolver.
Extended Eigensolver Functionality describes the problems that can and cannot be solved with the Extended Eigensolver and how to get the best results from the routines.
Extended Eigensolver Interfaces for eigenvalues within interval gives a reference for calling Extended Eigensolver routines.
Extended Eigensolver Interfaces for Extremal eigenvalues/singular values describes routines that can be used with Extended Eigensolver to find the largest/smallest eigenvalues or singular values.