Application

We offer Bachelor theses, Master theses, and student practicals in the research areas of the HAWAII Lab.

While not mandatory, we generally recommend completing a practical before starting a thesis. This gives students more time to become familiar with the research topic, tools, and workflow, and in our experience often leads to more successful thesis projects.

Applications should be sent to hawaii-thesis@uni-heidelberg.de.

What to include

To help us evaluate your background and identify suitable thesis topics, please include:

  • A short CV
  • A transcript of records (or another overview of your academic background). We are primarily interested in the courses you have completed, particularly those related to our research areas, as many thesis topics require knowledge from one or more of our lectures.
  • A brief motivation statement describing your research interests and the types of projects you would like to work on.

When to apply

We encourage students to apply well in advance of their intended start date. Finding a suitable topic and supervisor requires internal discussion, which typically takes several weeks. Contacting us early therefore increases the chances of finding a project that is a good match for both your interests and our current research activities.

Finding a topic

A good starting point for learning about our research are our lecture (see the Teaching page for an overview). You may also browse our Research Projects, our recent Publications and previously completed Theses to get an overview of the topics we work on.

We are happy to consider thesis topic proposals from students. However, our standard procedure is to match applicants against our internal list of available thesis topics based on their academic background, interests, and the prerequisites required for each project.

Completed Bachelor and Master Theses

Bachelor

  1. Jonas Kuhn, Parameterized Structured Pruning: Implementation and Experimental Analysis from Convolutional Networks to Vision Transformers, Bachelor of Science Computer Science, Heidelberg University, supervised by Holger Fröning, 2026
  2. Arjan Siddhpura, Studying the Feasibility of Object Detection on IPUs, Bachelor of Science Computer Science, Heidelberg University, supervised by Holger Fröning, 2025
  3. Marco Lorenz, Characterizing Detection Transformers for Hands, Guns and Phones with Roofline Methodology, Bachelor of Science Computer Science, Heidelberg University, supervised by Holger Fröning and E. W. Bethel, 2024
  4. Jonathan Bernhard, Optimized Bit-serial Operations on 64-bit ARM Processors using TVM Compilation and LIKWID Profiling, Bachelor of Science Computer Science, Heidelberg University, supervised by Holger Fröning and Stefan Riezler, 2023
  5. Florian Nowak, Instantiating the energy-based roofline model using hardware performance counters, Bachelor of Science Applied Computer Science, Heidelberg University, supervised by Holger Fröning, 2021
  6. Lisa Kuhn, Quantized Neural Networks for Keyword Spotting on Neuromorphic Hardware, Bachelor of Arts Computer Linguistics, Heidelberg University, supervised by Stefan Riezler and Holger Fröning, 2020
  7. Georg Weisert, CUDA Unified Memory - A deep dive, Bachelor of Science Applied Computer Science, Heidelberg University, supervised by Holger Fröning, 2020
  8. Raphael Kirchholtes, Data management for easy analysis of HPC communication traces, Bachelor of Science Applied Computer Science, Heidelberg University, supervised by Holger Fröning, 2020
  9. Otto von Zastrow-Marcks, Knowledge distillation for faster image segmentation, Bachelor of Science Applied Computer Science, Heidelberg University, supervised by Holger Fröning, 2020
  10. Matthias Hauck, A management environment for MEMSCALE, Bachelor of Science Computer Science, Heidelberg University, supervised by Holger Fröning, 2012

Master

  1. Dominik Gausepohl, Predicting LLM Training Energy Usage Based on Real World Measurements, Master of Science Computer Engineering, Heidelberg University, supervised by Holger Fröning and Ana Verbanescu, 2026
  2. Jonathan Bernhard, On Partial Bayesian Neural Networks from an MCMC Scalability Perspective, Master of Science Data and Computer Science, Heidelberg University, supervised by Holger Fröning and Jakob Zech, 2026
  3. Sun Shiejie, Profiling of Neural Networks, Master of Science Scientific Computing, Heidelberg University, supervised by Holger Fröning and Johannes Schemmel, 2026
  4. Jonathan Leis, Scheduling Optimizations for Distributed LLM Training, Master of Science Data and Computer Science, Heidelberg University, supervised by Holger Fröning and Ana Verbanescu, 2026
  5. Leandro Borzyk, Understanding LLM communication, Master of Science Computer Engineering, Heidelberg University, supervised by Holger Fröning and Dirk Koch, 2026
  6. Fangling Du, Bayesian Neural Networks for Robust Keyword Spotting with Uncertainty Quantification, Master of Science Scientific Computing, Heidelberg University, supervised by Holger Fröning and Jakob Zech, 2026
  7. Theo Stempel Hauburger, Sensitivity Metrics for Neural Networks in Noisy Environments, Master of Science Data and Computer Science, Heidelberg University, supervised by Holger Fröning and Johannes Schemmel, 2026
  8. Ilya Belkin, Efficient Fused Operators for Block Sparsity on GPUs, Master of Science Data and Computer Science, Heidelberg University, supervised by Holger Fröning and Jürgen Hesser, 2026
  9. Zhang Hao, BNN-KAN: Bayesian Neural Networks with Learnable Activation Functions for Enhanced Uncertainty Quantification, Master of Science Scientific Computing, Heidelberg University, supervised by Holger Fröning and Jürgen Hesser, 2026
  10. Max Mielke, Uncertainty Estimation for Singe Stage Object Detection, Master of Science Computer Engineering, Heidelberg University, supervised by Holger Fröning and Ullrich Köthe, 2025
  11. Paul Kupper, Exploiting Heterogeneous Computing for Stochastic Variational Inference based Bayesian Neural Networks, Master of Science Data and Computer Science, Heidelberg University, supervised by Holger Fröning and Dirk Koch, 2025
  12. Nils Kochendörfer, Exploring Memory Utilization on the IPU, Use Case: Needleman-Wunsch Algorithm, Master of Science Computer Engineering, Heidelberg University, supervised by Kazem and Holger Fröning, 2025
  13. Jiufeng Li, Quantization of activation during backward propagating, Master of Science Data and Computer Science, Heidelberg University, supervised by Holger Fröning and Robert Strzodka, 2025
  14. Constantin Nicolai, On Energy Modeling of Deep Neural Network Operations, Master of Science Computer Engineering, Heidelberg University, supervised by Holger Fröning and Robert Strzodka, 2025
  15. Anusha Chattopadhyay, Hyperparameter Optimization Algorithms for SLURM managed Compute Clusters, Master of Science Data and Computer Science, Heidelberg University, supervised by Holger Fröning and Vincent Heuveline, 2025
  16. Yong Wu, Multi-Level Quantization of Stochastic Variational Inference based Bayesian Neural Networks, Master of Science Data and Computer Science, Heidelberg University, supervised by Holger Fröning and Johannes Schemmel, 2025
  17. Christian Simonides, Efficient Ensemble-based Bayesian Neural Networks for Depth Regression, Master of Science Data and Computer Science, Heidelberg University, supervised by Holger Fröning and Franz Pernkopf, 2025
  18. Sergei Bespalov, Reducing Global Memory Accesses in DNN Training using Structured Weight Masking, Master of Science Computer Engineering, Heidelberg University, supervised by Holger Fröning and Dirk Koch, 2025
  19. Alper Daggez, Combined Sparsity and Quantization Compression for DNNs on FPGAs, Master of Science Data and Computer Science, Heidelberg University, supervised by Holger Fröning and Dirk Koch, 2025
  20. Tamara Bucher, Guided Galen: Guiding Automatic Compression of Neural Networks with Hardware Performance Metrics, Master of Science Computer Engineering, Heidelberg University, supervised by Holger Fröning, 2025
  21. Congcong Xu, Bayesian Bridge: Transferring Posterior Predictions Between Models Trained via Markov Chain Monte Carlo and Stochastic Variational Inference, Master of Science Scientific Computing, Heidelberg University, supervised by Holger Fröning and Gregor Schiele, 2025
  22. Prakriti Jain, Image Corruption on Object Detection Data to Evaluate Uncertainty Estimation, Master of Science Data and Computer Science, Heidelberg University, supervised by Holger Fröning and Ulrich Köthe, 2025
  23. Zhang Yi, Utilizing Machine Learning Methods for Time Series Classification, Master of Science Scientific Computing, Heidelberg University, supervised by Holger Fröning and Yufei Mao, 2024
  24. Xiao Wang, Improving the Robustness of DNNs to Noisy Computations, Master of Science Scientific Computing, Heidelberg University, supervised by Holger Fröning and Artur Andrzejak, 2024
  25. Hadi Ghaeni, Analysis of the Suitability of Federated Learning Approaches for Quality Data of Eroding Products, Master of Science Computer Engineering, Heidelberg University, supervised by Holger Fröning and Benjamin Kormann, 2024
  26. Lisa Kuhn, Scalability of Bayesian Neural Network Inference Methods for Real-World Tasks, Master of Science Data and Computer Science, Heidelberg University, supervised by Holger Fröning and Franz Pernkopf, 2023
  27. Eric Matthias Kern, Optimized Calibration for Analog Computations Targeting Deep Neural Networks on the Example of BrainScaleS-2, Master of Science Computer Engineering, Heidelberg University, supervised by Holger Fröning and Johannes Schemmel, 2023
  28. Christian Alles, On the performance of butterfly approximations on the GraphCore IPU, Master of Science Computer Engineering, Heidelberg University, supervised by Holger Fröning and Kazem Shekofteh, 2023
  29. Daniel Barley, Reducing the state of large-scale MLPs by compressing the backward pass, Master of Science Computer Engineering, Heidelberg University, supervised by Holger Fröning, 2023
  30. Tobias Richstein, Characterization and approximation of the backwards path of large-scale language models, Master of Science Data and Computer Science, Heidelberg University, supervised by Holger Fröning and Stefan Riezler, 2022
  31. Joachim Meyer, Compiler-assisted optimizations for data-parallel paradigms in hipSYCL, Master of Science Computer Engineering, Heidelberg University, supervised by Holger Fröning and Vincent Heuveline, 2021
  32. Royden Ezra Wagner, Parsing multiple characters of JSON per cycle on FPGAs, Master of Science Computer Engineering, Heidelberg University, supervised by Holger Fröning and Jonas Dann, 2021
  33. Chenyang Zhu, Comparing performance of GPU and FPGA accelerators using finite element methods, Master of Science Computer Engineering, Heidelberg University, supervised by Holger Fröning and Vincent Heuveline, 2021
  34. Hendrik Borras, Exploring structured sparsity within data-flow architecture on reconfigurable hardware, nan, Heidelberg University, supervised by Holger Fröning, 2021
  35. Florian Brunner, Designing Hardware-Efficient Convolutional Neural Networks via Reinforcement Learning Based Neural Architecture Search, nan, Heidelberg University, supervised by Holger Fröning, 2021
  36. Paul Bethge, Resource-efficient keyword spotting using quantized LSTMs on FPGAs, Master of Science Computer Engineering, Heidelberg University, supervised by Holger Fröning and Günther Schindler, 2020
  37. Dilan Canpolat, Performance modeling of multi-GPU Communication, nan, Heidelberg University, supervised by Holger Fröning, 2020
  38. David Marquant, Exploring the integration of libraries in automated multi-GPU compilation on the example of LU decomposition, Master of Science Computer Engineering, Heidelberg University, supervised by Holger Fröning, 2019
  39. Michael Harbarth, Compile-time performance modelling for GPGPU kernels using control-flow aware basic-block Analysis, Master of Science Computer Engineering, Heidelberg University, supervised by Holger Fröning, 2019
  40. Roland Wydra, Visual Odometry for VTOL, Master of Science Computer Engineering, Heidelberg University, supervised by Holger Fröning and Ulrich Brüning, 2019
  41. Himanshu Tiwari, Supporting and understanding binarized neural networks in Theano, Master of Science Applied Computer Science, Heidelberg University, supervised by Holger Fröning, 2018
  42. Antsa Andriamboavonjy, Evaluating correlations among prediction performance, data complexity, reduced precision, and sparsity of neural networks, Master of Science Applied Computer Science, Heidelberg University, supervised by Holger Fröning, 2018
  43. Armin Schäffer, Investigating the power saving potential for hierarchical interconnection networks, Master of Science Computer Engineering, Heidelberg University, supervised by Holger Fröning, 2018
  44. Andreas Melzer, Compressing sparseternary weight tensors, Master of Science Computer Engineering, Heidelberg University, supervised by Holger Fröning and Artur Andrzejak, 2018
  45. Sven Nobis, Design and evaluation of a communication technique that leverages heterogeneous memory in accelerated clusters, Master of Science Data and Computer Science, Heidelberg University, supervised by Holger Fröning and Robert Strzodka, 2018
  46. Dennis Rieber, Characterization of GPU communication, Master of Science Computer Engineering, Heidelberg University, supervised by Holger Fröning, 2017
  47. Klaus Naumann, Exploring high-level synthesis for reconfigurable logic to improve time for contemporary machine learning algorithms, Master of Science Computer Engineering, Heidelberg University, supervised by Holger Fröning and Andreas Kugel, 2017
  48. Julian Schwing, Dynamic code generation and execution of user defined logic within graph traversal algorithms, Master of Science Applied Computer Science, Heidelberg University, supervised by Holger Fröning and Michael Gertz, 2017
  49. Arthur Kuehlwein, Hash tables for unordered message matching on SIMT processors, Master of Science Applied Computer Science, Heidelberg University, supervised by Holger Fröning and Benjamin Klenk, 2017
  50. Steffen Lammel, Demonstrating energy saving potentials for high-performance interconnection networks using a power-aware network simulator, Master of Science Computer Engineering, Heidelberg University, supervised by Holger Fröning and Felix Zahn, 2017
  51. Lorenz Braun, Code feature supported automated partitioning and communication prediction for multi-GPU applications, Master of Science Computer Engineering, Heidelberg University, supervised by Holger Fröning, 2017
  52. Dominik Sterk, Optimized bulk data transfer in multi-GPU systems for improved total transfer time, Master of Science Applied Computer Science, Heidelberg University, supervised by Holger Fröning and Artur Andrzejak, 2016
  53. Günther Schindler, GPU architecture extensions for advanced communication and synchronization, Master of Science Computer Engineering, Heidelberg University, supervised by Holger Fröning, 2016
  54. Christoph Klein, Automated partitioning of data-parallel programs, Master of Science Physics, Heidelberg University, supervised by Ulrich Brüning and Holger Fröning, 2016
  55. Daniel Schlegel, Active messaging in autonomous GPU networks, Master of Science Computer Engineering, Heidelberg University, supervised by Holger Fröning and Robert Strzodka, 2016
  56. Eugen Rusakov, Performance monitoring and optimization for Theano-based deep learning on ARM processors, Master of Science Computer Engineering, Heidelberg University, supervised by Holger Fröning and Andreas Kugel, 2016
  57. Benjamin Baumann, A performance model for the training of DNNs on GPUs, Master of Science Computer Engineering, Heidelberg University, supervised by Holger Fröning and Robert Strzodka, 2016
  58. Julian Romera, Optimizing communication by compression for multi-GPU scalable breadth-first searches, Master of Science Computer Engineering, Heidelberg University, supervised by Holger Fröning, 2016
  59. Matthias Hauck, Scalable breadth first search using distributed GPUs, Master of Science Computer Engineering, Heidelberg University, supervised by Holger Fröning, 2014
  60. Alexander Matz, Extending MEMSCALE by an optimized integration into coherence domains, Master of Science Computer Engineering, Heidelberg University, supervised by Holger Fröning, 2013
  61. Benjamin Klenk, Comparing different communication paradigms for data-parallel processors, Master of Science Computer Engineering, Heidelberg University, supervised by Holger Fröning, 2013
  62. Elena Kuss, Analyzing power efficiency and cost effectiveness of direct and indirect interconnetion network topologies, Master of Science Computer Engineering, Heidelberg University, supervised by Holger Fröning, 2012