Dipl.-Inf. Andreas Schäfer
Since June 2009, Andreas Schäfer is a member of the research staff at the Chair of Computer Architecture. He holds a master's degree in computer science (Diplom-Informatiker).
Contact
| Address: |
Dipl.-Inf. Andreas Schäfer Lehrstuhl für Informatik 3 (Rechnerarchitektur) Martensstr. 3 91058 Erlangen Germany |
|---|---|
| Office: | 07.153 |
| Phone: | +49 9131 85 27910 |
| Fax: | +49 9131 85 27912 |
| Homepage: | http://www3.informatik.uni-erlangen.de/Persons/schaefer/ |
| E-Mail: | andreas.schaefer@informatik.uni-erlangen.de |
Research Interests
My research takes place at the boundary between scientific computing applications and high-performance computing: I provide other scientists with tools, enabling them to harness today's parallel computers. For my PhD thesis I focus on LibGeoDecomp, an auto-parallelizing library for stencil codes.
Apart from research, my position also includes teaching (e.g. the lecture Architectures of Supercomputers, which I hold in the winter terms) and administration of our chair's HPC servers.
Scientific Computing
- computer simulation
- stencil codes
- geometric decomposition codes (e.g. meshfree methods)
High-Performance Computing
- cluster computing (particularly multi-clusters)
- multi-cores (cache effects, vectorization)
- accelerators (GPGPUs, Cell BE, and FPGA integration)
Publications
- Schäfer, Andreas; Kronfeldner, Konstantin; Heller, Thomas:
A Scalable Backend for True MMORPGs -- Bringing Supercomputer Tech Back to Gaming.
In: Wiland, Ilka (Ed.): GPU Technology Conference -- Poster Track
(GPU Technology Conference, San Jose, 18.03.2013). vol. 1
2013, pp. 1-1. - Schäfer, Andreas; Reichenbach, Marc; Fey, Dietmar:
Continuous Integration and Automation for Devops.
In: Kim, Haeng Kon; Ao, Sio-Iong; Rieger, Burghard B. (Ed.): IAENG Transactions on Engineering Technologies Special Edition of the World Congress on Engineering and Computer Science 2011 1. ed.
Dordrecht: Springer, 2013, pp. 345-358. (Lecture Notes in Electrical Engineering) - ISBN 9400747853 - Schäfer, Andreas; Fey, Dietmar:
A Predictive Performance Model for Stencil Codes on Multicore CPUs.
In: Vuduc, Richard ; Iwashita, Takeshi (Ed.): The Seventh International Workshop on Automatic Performance Tuning
(The Seventh International Workshop on Automatic Performance Tuning, Kobe, 17.07.2012).
2012, pp. 1-13. - Schäfer, Andreas; Fey, Dietmar:
An Architecture Concept for the Scalable Simulation of Dendritic Growth.
Talk: VECPAR 2012, RIKEN Advanced Institute for Computational Science (AICS)
Kobe, 19.07.2012 - Schäfer, Andreas:
Ingredients for Ultra-Scalable Simulation Codes.
Talk: Facing the Multicore-Challenge III, Hochschule für Technik Stuttgar
Stuttgart, 19.09.2012 - Schäfer, Andreas:
LibGeoDecomp - A Library to Make Computer Simulations Scale/Interactive.
Talk: Seminar Institut M-8, Fluiddynamik und Schiffstheorie, Technische Universität Hamburg-Harburg
Hamburg, 14.12..2012 - Schäfer, Andreas:
LibGeoDecomp Developer Guide.
Talk: , FAU Erlangen
Erlangen, 11.01.2012 - Schäfer, Andreas:
Scaling a Reverse Time Migration Algorithm on the TSUBAME 2.0 Supercomputer.
Talk: Aoki Lab Meeting, Tokyo Institute of Technology
Tokio, 01.08.2012 - Schäfer, Andreas; Reichenbach, Marc; Fey, Dietmar:
Collaborative Administration in the Context of Research Computing Systems.
In: Ao, S.I. ; Douglas, Craig (Ed.): World Congress on Engineering and Computer Science 2011
(WCECS 2011, Berkeley, CA, USA, 19.-21.10.2011). vol. 1, 1. ed.
San Francisco: IAENG, 2011, pp. 1092-1097. - ISBN 9789881925176 - Seidler, Ralf; Schmidt, Michael; Schäfer, Andreas; Fey, Dietmar:
Efficient Implementation of Selected Parallel Path Planning Algorithms on GPUs.
In: Global Science and Technology Forum (GSTF); Bharat Bhargava: Journal on Computing vol. 1, 1. ed.
Singpur: Global Science and Technology Forum (GSTF), 2011, pp. 107-112. (GSTF International Journal on Computing (JoC) vol. 1) - Seidler, Ralf; Schäfer, Andreas; Fey, Dietmar:
Fast Dot Correlation in Optical Metrology on GPGPUs.
In: Arabni, Hamid R.; Uto, Minoru; Joe, Kazuki; Nishikawa, Hiroaki; Ishii, Hiroshi; Tinetti, Fernando G.; Solo, Ashu M. G.; Gravvanis, George A. (Ed.): Proceedings of The 2011 International Conference on Parallel and Distributed Processing Technology and Applications
(PDPTA 2011, Las Vegas, 18.Jul-21.Jul 2011). vol. 1, 1. ed.
Las Vegas: CSREA Press, 2011, pp. 248-254. - ISBN 1-60132-193-7 - Schäfer, Andreas; Fey, Dietmar:
High Performance Stencil Code Algorithms for GPGPUs.
In: Sato, Mitsuhisa ; Matsuoka, Satoshi ; Sloot, Peter M. ; van Albada, G. Dick ; Dongarra, Jack (Ed.): Proceedings of the International Conference on Computational Science
(ICCS 2011, Nanyang Technological University, Singapur, 01.-03.06.2011).
2011. - Schäfer, Andreas; Fey, Dietmar:
High Performance Stencil Code Algorithms for GPGPUs.
In: Sato, Mitsuhisa ; Matsuoka, Satoshi ; Sloot, Peter M. ; van Albada, G. Dick ; Dongarra, Jack (Ed.): Proceedings of the International Conference on Computational Science
(ICCS 2011, Nanyang Technological University, Singapur, 01.-03.06.2011).
2011. - Schäfer, Andreas:
LibGeoDecomp - How To Create Teraflop Simulations Within Minutes.
Talk: ASIM Workshop 2011, Leibniz-Rechenzentrum München
Garching, 15.03.2011 - Schäfer, Andreas; Hammer, Julian ; Fey, Dietmar:
Parallel Simulation of Dendritic Growth On Unstructured Grids.
Talk: SC11: Workshop on Irregular Applications: Irregular Applications: Architectures & Algorithms, SC11
Seattle, 13.11.2011 - Schäfer, Andreas; Fey, Dietmar; Hammer, Julian:
Parallel simulation of dendritic growth on unstructured grids.
In: Feo, John; Villa, Oreste ; Tumeo, Antonino ; Secchi, Simone (Ed.): IAAA '11: Proceedings of the first workshop on Irregular applications: architectures and algorithms
(SC'11 SC Conference on High Performance Computing Networking, Storage and Analysis Seattle, Seattle, 13.11.2011).
New York, NY, USA: ACM, 2011, pp. 15-22. - ISBN 978-1-4503-1121-2 - Schäfer, Andreas:
Past, Present and Future of LibGeoDecomp, an Auto-parallelizing Stencil Code Library.
Talk: Computing Sciences Seminar, Lawrence Berkeley Lab
Berkeley, CA, USA, 19.10.2011 - Seidler, Ralf; Schmidt, Michael; Schäfer, Andreas; Fey, Dietmar:
Comparison of Selected Parallel Path Planning Algorithms on GPGPUs and Multi-Core.
In: Global Science and Technology Forum (GSTF) (Org.): Proc. of the Annual International Conference on Advances in Distributed and Parallel Computing (ADPC 2010)
(Annual International Conference on Advances in Distributed and Parallel Computing (ADPC 2010), Singapur, 1.-2. November 2010).
Singapur: Global Science and Technology Forum (GSTF), 2010, pp. A133-A139. - ISBN 978-981-08-7656-2 - Schäfer, Andreas:
Geometrische Zerlegung.
In: Fey, Dietmar (Ed.): Grid-Computing: Grid Computing für Computational Science
Berlin: Springer, 2010, pp. 439-466. - ISBN 3540797467 - Schäfer, Andreas; Fey, Dietmar:
Tool for Automated Generation of MPI Typemaps.
In: Alameh, Kamal ; Sokolinsky, Leonid B. ; Chao-Tung, Yang (Ed.): International Conference on High Performance Computing Systems (HPCS-10)
(International Conference on High Performance Computing Systems (HPCS-10), Orlando, FL, USA, 12-14.07.2010).
Orlando: ISRST, 2010, pp. 92-96. - ISBN 978-1-60651-020-9 - Erdmann, Jakob; Schäfer, Andreas; Kauhaus, Andreas; Fey, Dietmar:
A profitability heuristic that reduces the parameter dependence of dynamic load balancing.
In: GI e.V. Bonn (Ed.): Informatiktage 2008 Fachwissenschaftlicher Informatik-Kongress
Bonn/Berlin: Köllen, 2008, pp. 217-219. (GI-Edition. Seminar 6 vol. LNI) - ISBN 978-3-88579-440-0 - Schäfer, Andreas; Fey, Dietmar:
LibGeoDecomp: A Grid-enabled Library for Geometric Decomposition Codes.
In: Lastovetsky, Alexey; Kechadi, Tahar; Dongarra, Jack (Ed.): Lecture Notes in Computer Science: Recent Advances in Parallel Virtual Machine and Message Passing Interface
(15th European PVM/MPI Users’ Group Meeting, Dublin, 2008). vol. 5205
Berlin: Springer, 2008, pp. 285-294. - ISBN 978-3-540-87474-4 - Schäfer, Andreas; Fey, Dietmar:
Pollarder: An architecture concept for self-adapting parallel applications in computational science.
In: Bubak, Marian; Albada, G. Dick van; Dongarra, Jack; Sloot, Peter M.A. (Ed.): Computational Science - ICCS 2008
Berlin, Heidelberg: Springer Verlag, 2008, pp. 174-183. (Computational Science - ICCS 2008 vol. LNCS) - ISBN 978-3-540-69383-3 - Schäfer, Andreas:
Job Distribution Framework for Large Scale Clusters and Grids.
In: IP.com Prior Art Database (2007), pp. 1-7 - Schäfer, Andreas; Erdmann, Jakob; Fey, Dietmar:
Simulation of Dendritic Growth for Materials Science in Multi-Cluster Environments.
In: Fey, Dietmar (Ed.): Workshop Grid-Technologie für den Entwurf technischer Systeme (Grid4TS)
(Workshop Grid-Technologie für den Entwurf technischer Systeme (Grid4TS),, Dresden, November 2007).
2007, pp. 1-8. - Schäfer, Andreas:
Harpy - High Throughput Computing.
Jena, Friedrich-Schiller-Universität Jena, Dipl. thesis, 2006 - Schäfer, Andreas:
Rock'n'Roll: An Engine for the Board Game "EinStein würfelt nicht".
Talk: Invited talk at the 17th Workshop in Discrete Optimization, Friedrich-Schiller-Universität Jena
Siegmundsburg, Germany, 2006 - Santosa, Mulyadi; Schäfer, Andreas:
Build a Heterogeneous Cluster with coLinux and openMosix.
In: IBM DeveloperWorks (2005), pp. 1-1 - Schäfer, Andreas; Fey, Dietmar:
Harpy - Employing a Process Migration Facility and a User Mode Linux for High Throughput Computing.
In: Gesellschaft für Informatik (Ed.): Informatiktage 2005: Fachwissenschaftlicher Informatik-Kongress
(GI Informatiktage 2005, Bonn, 8.-9. April 2005). 1. ed.
Bonn: Köllen, 2005, pp. 100-105. - ISBN 3885794365 - Kauhaus, Christian; Schäfer, Andreas; Fey, Dietmar:
Harpy: A Virtual Machine Based Approach to High-Throughput Cluster Computing.
Jena: Friedrich-Schiller-Universität Jena. 2005
- Internal Report - Schäfer, Andreas; Althöfer, Ingo:
Rock'n'Roll - A cross-platform engine for the board game "EinStein würfelt nicht".
Jena: Friedrich-Schiller-Universität Jena. 2005
- Internal Report - Schäfer, Andreas:
The "More for Less"-Paradox in Transportation Problems with Infinite-Dimensional Supply and Demand Vectors.
Talk: Kolloquium über Kombinatorik, Otto-von-Guericke-Universität Magdeburg
Magdeburg, 14.11..2004 - Althöfer, Ingo; Schäfer, Andreas:
The "More for Less"-Paradox in Transportation Problems with Infinite-Dimensional Supply and Demand Vectors.
Jena: Friedrich-Schiller-Universität Jena. 2004
- Internal Report
Screenshots of the Schülerinfotag 2013
Here you can see the screenshots we took during the Schülerinfotag 2013. The demo we ran paired our computer simulation library LibGeoDecomp with the parallel C++ runtime HPX and a particle-in-cell code. Instead of simulating protons in a fusion reactor, we did simulate brush strokes adapting to interactively changing force fields. The simulation and visualization ran on a GPU workstation, the live-steering was done from an Android tablet.
