We are happy to share that our journal article “A Systematic Evaluation of the Potential of Carbon-Aware Execution for Scientific Workflows” has been published in the journal Future Generation Computer Systems.

Scientific workflows are widely used for data-intensive scientific analysis and can result in significant energy consumption and carbon emissions when executed on clusters or cloud infrastructure. In this work, we evaluate the potential of different carbon-aware execution strategies using seven real-world Nextflow workflows across diverse computing environments, combining workflow execution traces with high-resolution grid carbon intensity (CI) data. Our results show substantial potential for emission reductions, with carbon-aware time shifting, for example, reducing footprints by over 80% using average CI data, and in some scenarios, completely using marginal signals. Meanwhile, using different processor governors for resource scaling can yield a footprint reduction of 67% under average CI.

This work significantly extends our earlier short paper presented at CCGrid 2025, providing a broader experimental evaluation and a more comprehensive analysis of carbon-aware execution strategies for scientific workflows.

The full journal article, co-authored with our close collaborators at HU Berlin, is available open access on ScienceDirect, and the evaluation code as well as experimental results are publicly available on GitHub.