Modeling DRAM in ACADL for Performance Predictions
Bachelor’s Thesis / Master’s Thesis / Student Research Project
Abstract
Abstract modeling of HW/SW systems is a relatively new research topic. This technique aims to capture only the essential parameters of software and hardware that influence their timing behavior.
On of the biggest influences on the performance and power of a computer architecture is the memory system where dynamic randon-access memory (DRAM) is almost always present.
This student project’s goal is to extend the Python-based Abstract Computer Architecture Description Language (ACADL) with a semantic for DRAM and apply existing performance prediction methods to evaluate the use of DRAM for different computer architectures and comparing the result against a cycle-accurate DRAM simulator such as DRAMSys.
References
- Konstantin Lübeck, Alexander Louis-Ferdinand Jung, Felix Wedlich, Oliver Bringmann - Work-in-Progress: Ultra-fast yet Accurate Performance Prediction for Deep Neural Network Accelerators
- Wikipedia - Dynamic random-access memory
Requirements
- Python
- Successfully atteded the lecture “Grundlagen der Rechnerarchitektur” and/or “Parallele Rechnerarchitekturen” (optional)
- Linux (optional)