Personal tools

Michael Rippl, M.Sc. (hons)

From Sccswiki

Jump to: navigation, search
Michael Rippl
Address:
TU München
Institut für Informatik
Boltzmannstr. 3
85748 Garching b. München
Office:
MI 02.05.059
Email:
Ripplmmail.png
Phone:
(089) 289 18 629
Fax:
(089) 289 18 607
Office hours:
by arrangement

Contents

Project

Eigenvalue Solver for Petaflop Applications: ELPA-AEO

  • Improvement of existing algorithms for the standard eigenvalue problem
  • Development of new algorithms for the generalized and standard eigenvalue problem


Research Interests

  • Numerical Linear Algebra
  • Efficient parallel eigenvalue computation
  • High Performance Computing


Possible Thesis topics

  • Performance Optimization of Numerical Linear Algebra code
  • Parallelization of Numerical Linear Algebra code
  • Developement and investigation of algorithms for Numerical Linear Algebra

Secretary of CSE

Responsibilities

  • Processing of Master's thesis
  • Questions and advice concerning examination regulations
  • Requests to the examination board

Teaching

Seminar Projektmanagement WiSe 17/18
Übung: Numerisches Programmieren SoSe 17
Übung: Numerisches Programmieren WiSe 16/17
Übung: Numerisches Programmieren SoSe 16

Talks

  • M. Rippl, T. Huckle and B. Lang: Efficient Transformation of the General Eigenproblem with Symmetric Banded Matrices to a Banded Standard Eigenproblem [BibTeX].
    PASC, Lugano, June 2017.
  • M. Rippl, T. Huckle and B. Lang: Efficient transformation of the general eigenproblem with symmetric banded matrices to a banded standard eigenproblem [BibTeX].
    SIAM CSE, Atlanta, March 2017.
  • M. Rippl: ELPA - Algorithmic extensions and optimization [BibTeX].
    PMAA, Bordeaux, July 2016.