To content

Static analysis for predicting the energy consumption of programs

-
in
  • News
  • Studium
das Bild zeigt einen gemalten, grünen Stromstecker und ein Blatt © Antonia van Ophuysen
The next lecture as part of the lecture series on computer science and sustainability will take place on Wednesday, June 4, 2025, 4:15 pm.

The next lecture as part of the lecture series on computer science and sustainability will take place on Wednesday, June 4, 2025, 4:15 p.m., in room E.23, Otto-Hahn-Str. 14.

Ben Hermann: Static analysis for predicting the energy consumption of programs

Abstract:

Measuring the power consumption of a program requires executing it. However, this is not always possible: there may be security concerns or perhaps the final execution environment is not yet fully defined. In addition, the measurement itself consumes energy. If you want to carry it out properly, you need to measure several runs. In my presentation, I will show you another way. Our SPEAR research project makes it possible to predict the energy consumption of any (C) program using only the program code. We use static program analysis methods and achieve results that are very close to actual measurements. This allows developers to pay attention to the energy consumption of their solution while programming.

Remarks:

The SPEAR project won the Sustainability Award of the Department of Computer Science 2024 in the Technical Sustainability category. It also won second prize in the Young Top-level Research Forum 2025: "Energy and Climate - Together for the Future" .

Further information