Research
Lowrank approximation and arithmetic
H-matrices, uniform H-matrices
standard, recursive arithmetic,
accumulator based arithmetic
lazy evaluation for lowrank arithmetic
task based scheduling for hierarchical lowrank arithmetic
detection of task dependencies in H-arithmetic
task frameworks for H-arithmetic
data compression
combination of lowrank and binary compression
adaptive compression modes
Projects
HLIBpro
parallel C++ library for H-matrix compression and arithmeticHLR
testbed for Hierarchical LowRank algorithmsHLIBcore
open-source, basic type and function set from HLIBproHLRcompress
lowrank approximation based data compression library
Publications
Articles
Ronald Kriemann, Hatem Ltaief, Minh Bau Luong, Francisco E. Hernández Pérez, Hong G. Im, David Keyes:
High-Performance Spatial Data Compression for Scientific Applications,
In: 28th International Conference on Parallel and Distributed Computing (Proceedings of EuroPar 2022), pp. 403–418, 2022Maseim Kenmoe, Ronald Kriemann, Matteo Smerlak and Anton Zadorin:
A fast iterative algorithm for near-diagonal eigenvalue problems,
SIAM Journal on Matrix Analysis and Applications, 43 4, pp. 1573-1598, 2022A. Litvinenko, R. Kriemann, M.G. Genton, Y. Sun and D. Keyes:
HLIBCov: parallel hierarchical matrix approximation of large covariance matrices and likelihoods with applications in parameter identification,
MethodsX, 7 (2020),S. Börm, S. Christophersen and R. Kriemann:
Semi-automatic task graph construction for H-matrix arithmetic,
SIAM Journal on Scientific Computing, Volume 44(2), pp. 77-98, 2022,N. Doucet, R. Kriemann, E. Gendron, D. Gratadour, H. Ltaief and D. Keyes:
Scalable soft real-time supervisor for tomographic AO,
In: Conference on adaptive optics systems VI : 10-15 June 2018, Austin, Texas, USA: SPIE, 2018 (Proceedings of SPIE ; 10703),José I. Aliaga, Rocío Carratalá-Sáez, Ronald Kriemann and Enrique S. Quintana-Ortí:
H-Task-parallel LU factorization of hierarchical matrices using OmpSs,
In: 2017 IEEE 31st international parallel and distributed processing symposium workshops (IPDPSW 2017); May 29 2017-June 2 2017 ; Orlando, Florida, USA New York : IEEE, 2017. - P. 1148-1157,Sabine Le Borne and Ronald Kriemann:
H-FAINV: Hierarchically factored approximate inverse preconditioners,
Computing and Visualization in Science, Volume 17, Issue 3, pp 135-150, 2015,Lars Grasedyck, Ronald Kriemann, Christian Löbbert, Arne Nägel, Gabriel Wittum, Konstantinos Xylouris:
Parallel tensor sampling in the hierarchical Tucker format,
Computing and Visualization in Science, Volume 17, Issue 2, pp 67-78, 2015,Ronald Kriemann:
H-LU Factorization on Many-Core Systems,
Computing and Visualization in Science, 16, pp. 105-117, 2015,Sabine Le Borne, Lars Grasedyck and Ronald Kriemann:
Domain-decomposition based H-LU preconditioning,
Numerische Mathematik, Vol. 112, pp. 565-600, 2009,Lars Grasedyck, Wolfgang Hackbusch and Ronald Kriemann:
Performance of H-LU Preconditioning for Sparse Matrices,
Computational Methods in Applied Mathematics, 8(4), pp. 336-349, 2008Lars Grasedyck, Ronald Kriemann and Sabine Le Borne:
Parallel blackbox H-LU preconditioning for elliptic boundary value problems,
Computing and Visualization in Sciences, Vol. 11, pp 273-291, 2007Sabine Le Borne, Lars Grasedyck and Ronald Kriemann:
Domain-decomposition based H-LU preconditioners,
Domain Decomposition Methods in Science and Engineering XVI, O.B. Widlund and D.E. Keyes (eds.), Springer LNCSE 55, pp. 667-674, 2007Mario Bebendorf, Ronald Kriemann:
Fast Parallel Solution of Boundary Integral Equations and Related Problems,
Computing and Visualization in Sciences, 8/3-4, pp. 121-135, 2005Wolfgang Hackbusch, Boris N. Khoromskij and Ronald Kriemann:
Direct Schur complement method by domain decomposition based on H-matrix approximation,
Computing and Visualization in Sciences, 8/3-4, pp. 179-188, 2005Wolfgang Hackbusch, Boris N. Khoromskij and Ronald Kriemann:
Direct Schur Complement Method by Hierarchical Matrix Techniques,
Domain Decomposition Methods in Science and Engineering, LNCSE, Vol. 40, pp. 581-589, 2004Ronald Kriemann:
Parallel H-Matrix Arithmetics on Shared Memory Systems,
Computing 74, pp. 273-297, 2005Wolfgang Hackbusch, Boris N. Khoromskij and Ronald Kriemann:
Hierarchical Matrices based on a Weak Admissibility Criterion,
Computing 73, pp. 207-243, 2004W. Hackbusch, R. Kriemann, S. Le Borne, J.-F. Maitre:
CD2D3D - a package to solve convection dominated problems employing ordering techniques,
Notes on Numerical Fluid Mechanics Vol. 75 (Springer), pp. 34-48, 2001.
Preprints
R. Kriemann:
Hierarchical Lowrank Arithmetic with Binary Compression
arXiv:2308.10960 [cs.MS]A. Litvinenko, R. Kriemann and V. Berikov:
Identification of unknown parameters and prediction with hierarchical matrices
arXiv:2104.07146 [stat.CO]
Talks
Winter School on Hierarchical Matrices 2024,
Parallel H-Matrices: Algorithms and Arithmetic,
Kiel University, Kiel, Germany,
2024-02-07Advanced numerical methods for non-local problems,
Reducing the Memory Gap between Hierarchical Lowrank Formats,
Boğaziçi University, Istanbul, Türkiye,
2024-01-0910th International Congress on Industrial and Applied Mathematics (ICIAM23),
Hierarchical Lowrank Arithmetic with Binary Compression,
Waseda University, Tokyo, Japan,
2023-08-22SIAM Conference on Computational Science and Engineering (CSE23),
Combining Binary Compression with Low-Rank Arithmetic,
Amsterdam, The Netherlands,
2023-03-0228th International European Conference on Parallel and Distributed Computing (EuroPar 2022),
High-Performance Spatial Data Compression for Scientific Applications,
University of Glasgow, Scotland, UK,
2022-08-25SIAM Conference on Parallel Processing for Scientific Computing (PP22),
Uniform-H, Bridging the Gap between H and H²,
Seattle, Washington, U.S.,
2022-02-23SIAM Conference on Computational Science and Engineering (CSE21),
Comparison of Low-Rank Update Techniques for H-Arithmetic,
2021-03-05SIAM Conference on Parallel Processing for Scientific Computing (PP20),
Semi-Automatic DAG Generation for (Tiled) H-Arithmetic,
Seattle, Washington, U.S.,
2020-02-13TC/PC² Kolloquium,
H-Matrices and H-Arithmetic on Many-Core-Systems,
University of Paderborn, Germany,
2018-09-10SIAM Conference on Applied Linear Algebra (ALA18),
Efficiency and Accuracy of Parallel Accumulator-based H-Arithmetic,
Hong Kong Baptist University,
2018-05-06SIAM Conference on Parallel Processing for Scientific Computing (PP18),
Accumulator based Task-parallel H-Factorization,
Waseda University, Tokyo, Japan,
2018-03-08European ExaScale Application Workshop,
Parallel H-Arithmetic On Many-Core Systems and Beyond,
School of Mathematics, University of Manchester, UK,
2016-10-11/12Scalable Hierarchical Algorithms for eXtreme Computing (SHAXC-3),
Parallel H-Arithmetic for Multi- and Many-Core Systems,
KAUST, Saudi Arabia,
2016-05-09.Colloquium of the Faculty of Engineering at the Dep. of Computer Science,
H-Arithmetic for Many-Core Systems,
University of Kiel, Germany,
2015-10-22H-Matrix Winterschool,
Task-Based H-Matrix Arithmetic, Part I: Algorithm Design,
Task-Based H-Matrix Arithmetic, Part II: Implementation,
MPI MIS Leipzig, Germany,
2014-02-27Numerische Behandlung von H-Matrizen mit der HLIBpro,
TU Bergakademie Freiberg, Germany,
2011-11-09Workshop “Complex Systems”,
Parallel Black Box H-LU Preconditioning,
METU Ankara, Turkey,
2009-05-15H-Matrix Winterschool,
H-Matrix Parallelisation,
Black Box Clustering and Parallel H-LU Factorisation,
MPI MIS Leipzig, Germany,
2009-03-04/05.GAMM/SIAM Conference on Applied Linear Algebra (ALA2006),
Parallel Hierarchical Matrices,
Universität Düsseldorf, Germany,
2006-07-2620th GAMM-Seminar Leipzig on “Numerical Methods for Non-Local Operators”,
Parallel H-matrix Arithmetic for Shared Memory Systems,
MPI MIS Leipzig,
2004-01-24.Workshop on “Algorithms for the Arithmetic of Dense Matrices from BEM/FEM”,
Parallel Algorithms for H-matrices,
Oberwolfach, Germany,
2001-07-12.
Lectures
Parallel Programming,
MPI MIS Leipzig, 2013,
Part 1: Introduction,
Part 2: Vectorisation,
Part 3: Shared Memory Programming,
Part 4: OpenMP,
Part 5: Threading Building Blocks,
LaTeX SourcesC++ for Scientific Computing,
MPI MIS Leipzig, 2008,
LaTeX Sources
Technical Reports
Ronald Kriemann:
HLIBpro C Language Interface,
Technical Report and Documentation 10/2008, MPI LeipzigRonald Kriemann:
HLIBpro User Manual,
Technical Report and Documentation 9/2008, MPI LeipzigRonald Kriemann:
Implementation and Usage of a Thread Pool based on POSIX Threads,
Technical Report and Documentation 2/2003, MPI Leipzig.Ronald Kriemann, Jens Burmeister and Rainer Kleinrensing:
Benchmarking a Shared Memory System,
Technical Report and Documentation 1/2003, MPI Leipzig
Theses
Ronald Kriemann:
Parallele Algorithmen für H-Matrizen,
PhD Thesis, Universität Kiel, 2005Ronald Kriemann:
Anordnungsalgorithmen für konvektionsdominante Probleme und deren Implementierung,
Diploma Thesis, Universität Kiel, 1999