Theses
Bachelor and Master theses are hosted and supervised by the research division at NHR@FAU, which is led by Prof. Dr. Harald Köstler (Department of Computer Science, FAU). The topics revolve around performance engineering, benchmarking, performance-aware application development, and tools development. We are happy to accept applications from all departments at FAU (contact us for details).
Open theses
- Bachelor/Master Thesis: Energy Modeling and Simulation of Parallel Applications on Clusters (contact: Ayesha Afzal)
- Bachelor/Master Thesis: Optimizing Parallel Programs using Persistent and Neighborhood Collectives (contact: Ayesha Afzal)
- Bachelor/Master Thesis: Development and in-core Performance Analysis with OSACA (contact: Jan Laukemann)
- Master Thesis: Performance Analysis of the SPEChpc 2021 Benchmark Suite (contact: Jan Laukemann)
Ongoing and finished theses
2024
- Nishtha Sharma: Physical Models for Symmetry Breaking in Distributed-Memory Parallel Programs. Master thesis (ongoing, advisors: A. Afzal, G. Hager)
- Simon Weber: Implementing frontend limitations in the in-core modeling tool OSACA. Bachelor thesis (ongoing, advisor: J. Laukemann)
- Krishna Sai Lakshmi Gayatri Manda: Performance Analysis of a Parallel Optical Flow Solver. Master thesis (ongoing, advisors: A. Afzal, G. Hager)
- Jean-Yves Verhaeghe: Exploring advanced visualization of MPI-parallel programs. Master thesis (ongoing, advisors: A. Afzal, G. Hager)
- Jonah Holtmann: Power consumption properties of modern GPUs. Bachelor thesis (ongoing, advisor: D. Ernst)
2024
- Michael Panzlaff: Optimizing sparse matrix-vector multiplication on AMD GPUs. Master thesis (advisor: D. Ernst)
- Aditya Ujeniya: Extending a Simulation Framework for Performance Assessment of Parallel Applications. Master thesis (advisor: A. Afzal)
- Jairo Andres Buitrago Franco: Exploring Distributed-Memory Parallelization Algorithms for Molecular-Dynamics Simulations in MD-Bench. Master thesis (advisor: J. Eitzinger)
- Stefan de Souza: Power consumption properties of modern multicore server CPUs. Master thesis (advisors: G. Hager and V. Aizinger [University of Bayreuth])
- Gonzalo Pinzon Walteros: Unveiling the tricks and optimizations of the HPCG benchmark using performance analysis. Master thesis (advisor: C. L. Alappat)
2023
- Arsenii Andropov: Implementing production-grade MD algorithms in MD-Bench. Master thesis (advisor: R. Ravedutti Lucio Machado)
- Dane C. Lacey: A Novel Cache-Blocked MPI-parallel Matrix Power Kernel: Application to Finite-Volume Methods in Cardiac Arrhythmia Simulations. Master thesis (advisor: C. L. Alappat)
2022
- Julian Hammer: Design and Implementation of an Automated Performance Modeling Toolkit for Regular Loop Kernels. Dissertation, FAU Erlangen-Nürnberg
2021
- Achanta Ravi Chandra: Analysis of Single-Channel Patch-Clamp Time Series using a 2D-Fit Algorithm and Performance Evaluation on HPC systems. Master thesis (advisor: T. Gruber).
- Sebastian Bönning: Integration of chip-level performance models into a parallel simulation framework. Bachelor thesis (advisors: A. Afzal, G. Hager).
- Matthias Knorr: Stencil algorithms on GPGPUs. Master thesis (advisor: D. Ernst).
2020
- Dorsaf Jdidi: Classification of HPC jobs using unsupervised Machine Learning on job performance metric timeseries data. Bachelor Thesis (advisor: J. Eitzinger in collaboration with the Chair of Pattern Recognition at FAU)
- Linda Stadter: Classification of HPC jobs using supervised Machine Learning on job performance metric timeseries data. Bachelor Thesis (advisor: J. Eitzinger in collaboration with the Chair of Pattern Recognition at FAU)
- Jan Laukemann: Cross-Architecture Automatic Critical Path Detection For In-Core Performance Analysis. Master thesis (advisor: J. Hammer)
2019
- Hassan Asghar: Development of a benchmark suite for investigating MPI communication behavior. Master thesis (advisors: A. Afzal, G. Hager).
- Sebastian Kuckuk: Automatic Code Generation for Massively Parallel Applications in Computational Fluid Dynamics. Dissertation, FAU Erlangen-Nürnberg (advisor: H. Köstler)
2018
- Thomas Köster: Porting Physical Parameterizations from a Climate Model to Accelerators Using OpenACC Compiler Directives. Master thesis (cooperation with Università della Svizzera Italiana, Lugano; advisors: G. Wellein, O. Schenk).
- Michael Hußnätter: Performance Engineering for Lattice-Boltzmann Kernels on an Intel Many-Core Architecture with Wide SIMD Units and High-Bandwidth Memory. Master thesis (advisor: M. Wittmann).
2017
- Jan Laukemann: Design and Implementation of a Framework for Predicting Instruction Throughput. Bachelor thesis (advisors: J. Hammer, J. Hofmann).
- Karnajit Sen: A Technical Assessment and Micro-Benchmarking of the NVLink Interconnect for IBM Power8 Systems with NVIDIA GPUs. Master thesis together with IBM Deutschland R&D GmbH (advisors: M. Kreutzer, G. Wellein).
- Dominik Ernst: Performance Engineering for Block Vector Matrix Multiplications on GPUs. Master thesis (advisor: M. Kreutzer).
- Moritz Kreutzer: Performance Engineering for Exascale-Enabled Sparse Linear Algebra Building Blocks. Dissertation, FAU Erlangen-Nürnberg (2017).
2016
- Markus Wittmann: Hardware-effiziente, hochparallele Implementierungen von Lattice-Boltzmann-Verfahren für komplexe Geometrien. Dissertation, FAU Erlangen-Nürnberg (2016).
- Christie L. Alappat: Implementation and Performance Engineering of the Kaczmarz Method for Parallel Systems. Master thesis (advisors: G. Wellein, M. Kreutzer, J. Thies [DLR Köln]).
- Michael Hußnätter: Performance Analysis and Optimization of Indirect Addressing Lattice Boltzmann Schemes in the waLBerla Framework. Bachelor thesis (advisors: U. Rüde [LSS], C. Godenschwager [LSS], M. Bauer [LSS], M. Wittmann).
- Dominik Thoennes: Implementation options and performance properties of OpenMP-parallel symmetric sparse matrix-vector multiplication. Master thesis (advisor: G. Hager).
2015
- Ayesha Afzal: The Cost of Computation: Metrics and Models for Modern Multicore-based Systems in Scientific Computing. Master thesis (advisor: G. Hager)
- Salah Saleh: Performance Analysis and Modeling of Computational Kernels for the Simulation of Signal Transmission in the Brain. Master thesis (advisor: G. Hager)
- Julian Hammer: Automatic Loop Kernel Analysis and Performance Modeling. Master thesis (advisor: G. Hager)
- Dominik Ernst: Modeling and optimization of a stencil kernel on the Intel Xeon Phi coprocessor. Bachelor thesis (advisor: G. Hager)
- Johannes Habich: A performance engineering process for developing high performance Lattice Boltzmann implementations. Dissertation, FAU Erlangen-Nürnberg (2015)
2014
- Yi Zhou: Implementation of fast sorting and selection algorithms on modern architectures. Master thesis (advisor: J. Eitzinger)
2013
- Johannes Hofmann: Performance Evaluation of the Intel Many Integrated Core Architecture for 3D Image Reconstruction in Computed Tomography. Master thesis (advisors: G. Wellein, J. Treibig)
- Johannes Bleisteiner: Optimization strategies for a Lattice Boltzmann Method with sparse data representation on the Intel Xeon Phi co-processor. Master thesis (advisor: G. Hager)
2012
- Tobias Scharpff: Analyse und Optimierung von Operationen auf dünn besetzten Matrizen. Studienarbeit, betreut von G. Hager und K. Iglberger (LSS)
- Klaus Sembritzki: Evaluation of the Coarray Fortran Programming Model on the Example of a Lattice Boltzmann Code. Master thesis (advisors: G. Wellein, J. Treibig, B. Krammer [Université de Versailles St-Quentin-en-Yvelines])
2011
- Nan Chen: Comparative Study of OpenFOAM and ANSYS CFX. Master thesis (advisors: T. Zeiser, G. Zheng)
- Faisal Shahzad: Checkpoint/Restart for Fault Tolerant MPI Programs: A Case Study using a Lattice Boltzmann Code and the NAS Parallel Benchmarks. Master thesis (advisor: G. Wellein)
2010
- Holger Stengel: Paralleles Programmieren auf hybrider Hardware: Modelle und Anwendungen. Masterarbeit, betreut von G. Hager und R. Eck (Georg-Simon-Ohm Hochschule Nürnberg)
2009
- Markus Wittmann: Potentials of temporal blocking for stencil-based computations on multi-core systems. Master thesis (advisors: G. Hager and R. Eck [University of Applied Sciences Nuremberg])
2008
- Johannes Habich: Performance Evaluation of Numeric Compute Kernels on nVIDIA GPUs. Master thesis (advisors: G Wellein and U. Rüde [LSS])
2007
- Holger Stengel: C++-Programmiertechniken für High Performance Computing auf Systemen mit nichteinheitlichem Speicherzugriff unter Verwendung von OpenMP. Diplomarbeit, betreut von G. Hager (RRZE) und R. Eck (FH Nürnberg)
2006
- Johannes Habich: Improving computational efficiency of lattice Boltzmann methods on complex geometries. Bachelor Thesis (RRZE, LSS)
- Aditya Nitsure: Implementation and Optimization of a Cache Oblivious Lattice Boltzmann Algorithm. Master thesis (RRZE, LSS)
2005
- Guillermo Zschaeck: On advanced lattice Boltzmann methods: analysis of different boundary conditions and evaluation of a novel hybrid thermal model. Master thesis (RRZE, LSTM)
2004
- Stefan Donath: On Optimized Implementations of the Lattice Boltzmann Method on Contemporary High Performance Architectures. Bachelor thesis
- Milena G. Gergova: Numerical Pore-Scale Studies of Transport Phenomena in Fixed-Bed Reactors. Master thesis (RRZE, CRT)
- Jürgen Bauer: Ortsaufgelöste Simulation von Transportprozessen in porösen Medien zur Untersuchung von Dispersion und Verweilzeitverhalten. Diplomarbeit (LSTM, RRZE, CRT)