SC20 Proceedings

The International Conference for High Performance Computing, Networking, Storage, and Analysis

Demystifying the Dark Arts of HPC: Introducing Biomedical Researchers to Supercomputers


Authors: Andrea Townsend-Nicholson, David Gregory, and Art Hoti (University College London (UCL), Division of Biosciences); Cristin Merritt and Stu Franks (Alces Flight Limited); Andrew Narracott (University of Sheffield; Department of Infection, Immunity and Cardiovascular Disease); Guillaume Hautbergue (University of Sheffield, Department of Neuroscience); Marcellus Augustine (University College London (UCL), University College London (UCL) Medical School); and Abdullah Ahmad, Joshua Palmer, Ailsa Jones, and Aaisha Sair (University of Sheffield, Medical School)

Abstract: The ability to bring personalised medicine from science fiction to science fact and into everyday clinical practice is the focus of the CompBioMed Centre of Excellence, a group of private, public and research organizations led by University College London. To reach this goal, CompBioMedā€™s Education, Training and Sustainability team teaches university students studying a broad range of disciplines to run biomedical and bioscience applications on high performance computing (HPC) resources. For this, students acquire skills in core computing methodologies, learning the "dark arts" of optimization to produce efficient, effective workloads. CompBioMed has created a scalable training model for remotely-delivered dark arts teaching, starting with optimization of the QIIME2 application on a cloud cluster they have named nUCLeus. This talk covers the collaborative efforts that CompBioMed has used to develop a repeatable scalable training model, and how students around the world can successfully learn to use HPC without breaking the bank.

Extended Abstract: pdf
Presentation: pdf



Back to the Visualization & Data Analytics Showcase Archive Listing