How did you hear about Green Compute for Research @UoM?
I heard about Green Compute for Research @UoM from Prof Caterina Doglioni (Dept. Physics and Astronomy) as part of our now completed project on the carbon costs of computing at the University. Within the carbon costs for computing project, I provided a hands-on approach with computer hardware for our PDRA to build their knowledge. My focus on hardware, combined with a background in climate science gives me a long-term perspective when discussing compute and I bring this to the Green Compute team.
What is your role at the University and how does it interact with the user group?
I’m a Senior Lecturer in Climate Resilience and project lead on the carbon efficient computing project. This project has led to a follow-on piece of work in the Network for sustainable Digital Research Infrastructure Vision and Expertise (NetDRIVE) programme where we’re focusing on training people to use computers in a more sustainable manner. Within Green Compute for Research @UoM we end up cross posting as a lot of the resources we find are relevant to both groups.
How did you get into your domain?
I graduated from UoM in Physics with Theoretical Physics before a PhD in Leeds in Climate Science. From there I moved to Paris and drifted into climate impacts, since returning to the UK, I’m more practical in my outlook and I’m interested in things we can change.
What does your role in the group involve?
I sit on the AI Environmental Sustainability working group with other group leads. My background of interest is in consumer grade hardware and as we diversify away from AI this is useful contextual knowledge.
What would you like to see the group doing over the next 12 months?
I’d like to see us have some strong practical advice for budget holders on what equipment to buy. Extending the lifespan of compute for teaching, learning and research, especially if we’re able to undervolt and overclock will save vast amounts of carbon. It would be great to work together with Research IT Services on this.
What do you hope to come out the group?
I’d like green computing to become a default behaviour as common as recycling. Where we’ve found good examples, we share them in the Green Compute @UoM channel.
How can members of the group get involved?
Join the group, let us know what they’re working on. We’re interested in all areas of good practice, so if people want to lead or attend workshops, demonstrate efficiency gains in code, or share links to external training, then let us know.
What do you enjoy most about the group?
I like seeing people trying to solve the problem and in many cases succeeding in making their compute work greener. These initial steps are small, but we can also see people getting into the habit of asking the question, “Do I need to run this?”.
Fun fact about you?
I’ve danced in the ballroom of what at the time felt like a mostly abandoned ocean liner.