Skip to content. | Skip to navigation

Personal tools

Navigation

Home > ELPA goals

ELPA goals

Main goal is to develop efficient algorithms for the eigenproblem of large symmetric matrices  and to implement, parallelize, port and optimize them for the usage on very large processor core numbers of PetaFlop computers in such a way, that several scientific application areas can benefit. In particular the complete (or a big part of the whole) eigenspectrum has to be calculated.

From the point of view of the applied sciences involved a dominant goal is to enable the addressing of a new and highly relevant class of problems from Materials Science (electronic structure calculations) and from Technical and Biological Networks (structure analysis) as important scientific fields with needs for next generation supercomputers. Further disciplines are expected to benefit in multiple ways from such approaches for scalable eigenvalue solvers: Structural Mechanics (construction analysis), Fluid Mechanics (non-steady flows, complex turbulence models, non-linear optimization procedures for sound propagation), or Pattern Recognition.

Navigation
Log in


Forgot your password?