A centralised green computing course for digital researchers

Abstract:

Digital research covers a wide range of disciplines, and researchers have very different levels of technical experience. When we started designing a course on digital research sustainability, one of our main challenges was making sure it would be relevant to people from different backgrounds. 

To address this, we developed four personas (case studies) that represent common digital research roles: a Research Software Engineer, a computational lab researcher, an HPC user, and a machine learning practitioner. The first part of the course introduces key concepts and terminologies for green computing, while the second part uses these case studies to encourage participants to reflect on how the ideas apply to their own work. 

The iterations of the course that we have run so far, have attracted diverse audiences, from computational researchers to sustainability professionals. Their feedback has helped to further improve the course materials. As a positive outcome of developing this course, Imperial’s sustainability team have also launched a university-wide seven steps to ICT sustainability. 

In this talk, I will share some of the experiences the team had while developing and delivering this course. The audience will have the opportunity to reflect on and provide feedback on the topic. Some questions to consider would include: What are the pros and cons of adapting such a course to different research groups or disciplines under a wider institutional setting? What other approaches could RSEng teams take when (developing and) delivering digital research sustainability training? When considering the impact of digital research, should we stick to accounting for the carbon emissions only or also focus on other types of environmental impacts as well (e.g. water)?