Cloud HPC for a Living Laboratory
The Department of Biological and Environmental Sciences at the University of Stirling is transitioning satellite data modelling focused on water quality of the Firth of Forth into High Performance Computing (HPC) as part of a project to inform and help develop plans for a green monitoring platform for Scotland’s International Environment Centre (SIEC).
With an aim to deliver on the Scottish government’s commitment to Net Zero, the Department of Biological and Environmental Sciences at the University of Stirling are taking two decades of development in remote sensing technology and transforming it to meet the needs of the Forth Valley’s Green Recovery Programme. To scale and automate their Machine Learning tool sets, the team looked to explore the advantages presented by cloud-based HPC.
“There are currently two factors to consider when you want to get a complete environmental picture of an area,” says Dr. Peter Hunter, Senior Lecturer at the University of Stirling, “What you see on the ground and what you can detect from the space around and above. Our team is focused on the latter, with a specific aim to transform what we do into a flexible, scalable HPC process. Together, with the related data and workloads, we hope to create a robust investigative toolset for the Forth Valley.”
With the goal of looking at environmental impact, the University of Stirling looked to start this transition on public cloud, choosing the Microsoft Azure platform as a means to build and test the resources that could potentially make up the final solution. Their cloud HPC environment, built through a partnership with Phoenix Software and Alces Flight, features a persistent infrastructure core containing high-performance networking and storage along with on-demand compute resources. Jobs can be scheduled to run on right-sized AMD EPYC or Intel Xeon powered compute instances launched dynamically and accessed via the Slurm batch-job scheduler.
To read more about the University of Stirling’s project, including how this project is helping to transition more than just their piece of the project to HPC, read the complete story here(no registration required).