A 3D-Parallel Interior Eigenvalue Solver

Kurzfassung

Some applications in quantum physics require the computation of a relatively large part of the interior of the spectrum of the Hamiltonian matrix. The sparse matrices in question have a dimension of billions to trillions and on the order of 1000 eigenpairs are required. We present a purely iterative solver based on the FEAST algorithm (Polizzi '09) with a fault-tolerant and hybrid-parallel row-projection method for the linear systems that have to be solved. The subspace is distributed in both the `horizontal' and `vertical' direction, and the key operations exploit any available intra-node parallelism.