Anna Blume Jakobsen

Distributed software, such as cloud computing and internet applications, is a major part of much of the software that the modern world relies on. Unfortunately, distributed programs consume a lot of electrical power, which must be reduced for the sake of sustainability. This is hard to do with current science, because distributed programs are complex and their energy use is impacted by different factors.

Our project aims to address this challenge by providing, first, a rigorous mathematical model to reason about the energy use of distributed programs and, second, proven ways to rewrite such programs to lower energy consumption. Those tools can then be incorporated into programming editors to suggest energy optimizations to the programmer. We will ensure that every step is mathematically proven correct.

By offering a tractable method to reason about the energy costs of distributed programs, this project will be of scientific interest to computer scientists, engineers, and programmers. It will also show that we can empower programmers to reason about energy use without the need for expert knowledge. Since it will no longer be impractical to reason about energy efficiency, regulators will be enabled to impose restrictions on the energy use and sustainability of distributed software, which can have a positive impact on our planet.

‍