作者: Carolin Penke , Andreas Marek , Christian Vorwerk , Claudia Draxl , Peter Benner
DOI: 10.1016/J.PARCO.2020.102639
关键词:
摘要: We present a high-performance solver for dense skew-symmetric matrix eigenvalue problems. Our work is motivated by applications in computational quantum physics, where one solution approach to solve the so-called Bethe-Salpeter equation involves of large, dense, problem. The computed eigenpairs can be used compute optical absorption spectrum molecules and crystalline systems. One state-of-the art package symmetric matrices ELPA (Eigenvalue SoLvers Petascale Applications) library. extend methods available matrices. This way, presented method benefit from optimizations that make it well-established, efficient scalable library, such as GPU support. compare performance scalability our only matrices, an indirect route involving complex arithmetic. In total, we achieve up 3.67 higher than reference using Intel's ScaLAPACK implementation. runtime Bethe-Salpeter-Eigenvalue problem improved factor 10. freely current release