symmetric-tridiagonal-eigensolver
The final project for CS 51501 Parallelism in Numerical Linear Algebra during the spring semester in 2016 at Purdue University, it is a Fortran 90 implementation of a parallel eigensolver for symmetric tridiagonal matrices with Intel MKL, OpenMP and MPI. The algorithm uses Sturm sequence to compute inertias of the matrix with different diagonal shifts and then computes eigenvectors with inverse iteration.