Statistical Performance Estimation and Extrapolation
Bachelor’s Thesis / Master’s Thesis / Student Research Project
Abstract
This student project’s goal is to use statistical methods on benchmark data for Performance Estimation of Embedded AI Accelerator Architectures. The approach should be compard to state-of-the-art statistical performance modeling methods like ANNETTE or the Performance Representatives (PR).
References
- A. L.-F. Jung, J. Steinmetz, J. Gietz, K. Lübeck, und O. Bringmann, “It’s all about PR – Smart Benchmarking AI Accelerators using Performance Representatives”. arXiv, 12. Juni 2024.
- M. Wess, M. Ivanov, C. Unger, A. Nookala, A. Wendt, und A. Jantsch, “ANNETTE: Accurate Neural Network Execution Time Estimation With Stacked Models”
- AutoML for Tabular Data
Requirements
- Python
- Machine Learning
- Linux
- Successfully atteded the lecture “Grundlagen der Rechnerarchitektur”
- Successfully atteded the lecture “Modellierung und Analyse Eingebetteter Systeme” (optional)
- Successfully atteded the lecture “Efficient Machine Learning in Hardware” (optional)