I’m a graduate research assistant working on latency prediction models for AI workloads executed on accelerator hardware enabling fast optimization loops for neural network architecture search (NAS) and pre-silicon hardware evaluation.
My expertise includes:
- 5+ years of Verilog programming and FPGA synthesis ⚙️⚡
- 4+ years in Deep Neural Network applications 🤖
- 4+ years of C/C++ programming for high-performance computing and embedded devices 🖥️
- 4+ years of Python programming for scientific analysis 🐍🧪
My colleagues describe me as a detail-oriented analytical thinker with a passion for learning new things. I am a team player because I enjoy tackling challenges as a group and discussing problem solutions from different angles to learn from others and share my knowledge.
I am currently working on my PhD thesis at the embedded systems group (Prof. Bringmann) at the University of Tübingen.
Publications
C. Gerum, A. Frischknecht, T. Hald, P. P. Bernardo, K. Lübeck, and O. Bringmann, “Hardware Accelerator and Neural Network Co-Optimization for Ultra-Low-Power Audio Processing Devices,” in 25th Euromicro Conference on Digital System Design (DSD), pp. 365–369, Maspalomas, 2022. PDF
K. Lübeck, A. L.-F. Jung, F. Wedlich, and O. Bringmann, “Work-in-Progress: Ultra-fast yet Accurate Performance Prediction for Deep Neural Network Accelerators,” in ACM/IEEE International Conference on Compilers, Architectures, and Synthesis for Embedded Systems (CASES), Shanghai, 2022. PDF
C. Grimm, F. Wawrzik, A. L.-F. Jung, K. Lübeck, S. Post, J. Koch, and O. Bringmann, “APPEL - AGILA ProPErty and Dependency Description Language,” in MBMV 2021: 24. GI/ITG/GMM- Workshop “Methoden und Beschreibungsprachen zur Modellierung und Verifikation von Schaltungen und Systemen”, München, 2021. PDF
P. P. Bernardo, C. Gerum, A. Frischknecht, K. Lübeck, and O. Bringmann, “UltraTrail: A Configurable Ultra-Low Power TC-ResNet AI Accelerator for Efficient Keyword Spotting,” in IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems (TCAD), pages 1-12, 2020. PDF
K. Lübeck and O. Bringmann, “A Heterogeneous and Reconfigurable Embedded Architecture for Energy-Efficient Execution of Convolutional Neural Networks,” in ARCS 2019: Architecture of Computing Systems, Copenhagen, 2019.
K. Lübeck, D. Morgenstern, T. Schweizer, D. Peterson, W. Rosenstiel, and O. Bringmann, “Neues Konzept zur Steigerung der Zuverlässigkeit einer ARM-basierten Prozessorarchitektur unter Verwendung eines CGRAs,” in MBMV 2016: 19. GI/ITG/GMM- Workshop “Methoden und Beschreibungsprachen zur Modellierung und Verifikation von Schaltungen und Systemen”, Freiburg, 2016. PDF