PRACE Summer of HPC 2021

It was apparent from the start that travel would not be possible, and so this PRACE SoHPC for 2021 became a virtual event. Regrettably, this means that students do not get to travel to meet other programme participants either during the training week or at the hosting site. However two students instead of the usual one have been allocated to each project, which includes the aspect of remotely working in a team.

EPCC is managing three projects with a total of six students. These are:

• Investigating Scalability and Performance of MPAS Atmosphere Model managed by Dr Evgenij Belikov, EPCC.
  
  This is an investigation of the performance and scalability of the Model for Prediction Across Scales (MPAS) Atmosphere Model on ARCHER2 which will determine the performance characteristics of this code and, if there is time, attempt to perform some optimisation of the code. The project is being undertaken by Jonas Alexander Eschenfelder, who is studying Geophysics at Imperial College in London, and Carla Nicolin Schoder, who is currently finishing a Masters in Astrophysics at the University of Vienna.

• Re-engineering and optimizing Software for the discovery of gene sets related to disease.
  
  The project is looking to understand and improve the performance of the genomicper R package available through CRAN that performs pathway analyses to find associations between genes and disease. The aim is to increase the size of data sets that the package can handle and to be able to process existing data sets faster. This project is being undertaken by İrem Okur, a third-year Computer Engineering student at Dokuz Eylül University, Turkey, and Aybüke Özçelik, a graduate student in the Electronics and Communication Engineering Department at Izmir Institute of Technology, Turkey. EPCC's Dr Mario Antonioletti will manage the project, assisted by Dr Pau Navarro from the Institute of Genetics and Cancer at the University of Edinburgh, and Dr Claudia Cabrera from the Centre for Translational Bioinformatics at QMUL. 

• Performance of Parallel Python Programs on ARCHER2.
  
  Managed by Dr David Henty from EPCC, this project is investigating and improving the performance of a Computational Fluid dynamics of fluid flow in a cavity developed by a SoHPC project from last year. It is being undertaken by Alejandro Dinkelberg, who is currently doing a PhD at the University of Limerick combining modelling, network analysis and social psychology, and Jiahua Zhao, who is currently studying for a Masters degree in Geophysics at the University of Padova.

The PRACE SoHPC Programme gives late-stage undergraduates and Masters students a chance to experience a certain level of research undertaken at one of the top supercomputing centres in Europe. The Programme runs between July and August each year so, at the time of writing, the students are half way through their assignments. This year there are 66 students distributed across 33 projects. Despite the difficult circumstances we hope that the students will learn from their experience and have a good time in the process.