How to reduce the carbon footprint of your research computing

Speaker: Dr Nick Souter, School of Psychology, University of Sussex, UK

Abstract:

The storage and processing of large datasets, including neuroimaging data, uses energy, and therefore has a carbon footprint. In this 1-hour workshop, we’ll discuss how computing leads to carbon emissions, and what can be done to reduce your personal research computing footprint. Specifically, we’ll provide context on:

• The scale of impact of the ICT sector on the climate
• The impact of accumulating unused ‘dark data’
• The importance of carbon intensity
• How to cut back on unnecessary computing
• Current methods for tracking computing energy usage and carbon emissions

The content discussed here will include specific examples from neuroimaging research, including how one can reduce the carbon footprint of preprocessing in fMRIPrep, and the effect of fMRI software choice on energy usage. The messages and approaches discussed here will also apply to any discipline requiring the processing of large amounts of data, beyond neuroimaging alone. The session will end with live demonstrations of a digital carbon tracking tool and a climate aware task scheduler.