MMES Graduate Student Seminar: ELSI -- A Software Infrastructure for Electronic Structure Solvers
Computer simulations based on electronic structure theory, particularly Kohn-Sham density-functional theory (KS-DFT), are facilitating scientific discoveries across a broad range of disciplines such as chemistry, physics, and materials science. Despite its remarkable success, the application of KS-DFT to systems consisting of thousands of atoms is still difficult. The major computational bottleneck is an eigenvalue problem, whose direct solution scales cubically with respect to the problem size. To overcome this bottleneck, researchers and software developers are actively working on improving the performance of eigensolvers and developing alternative algorithms that circumvent the explicit solution of the eigenproblem. In this talk, we present ELSI, an open-source software infrastructure aiming to facilitate large-scale electronic structure calculations. ELSI features a unified programming interface that connects electronic structure codes to a variety of high-performance solvers. The efficiency and scalability of these solvers are demonstrated by a set of benchmarks performed on world-leading supercomputers.