Accelerating your research using attractive parallel programming and HPC
The procedures in research and time to investigate to get a solid work done — surprisingly needs patience and time. Therefore, you try to turn every screw to improve some processes. For me, getting faster results (good or bad) in the simulations we do is one way to tackle it. Besides improving code if you run the same loop or simulation over and over again, parallelize it and distribute the tasks is a promising path!
Last year I started as a PhD-Candidate in the DAFINET-team at the University of Limerick to combine modelling, network analysis and social psychology to get an insight of how attitudes are represented in us and how they are lead by our salient social identities. My part of role is to back up new experiments by simulations, mainly in Python.
I started to look out for learning opportunities to improve coding and their speed of delivering results. I ran into the homepage of PRACE, which offers various training events for my needs (and maybe yours as well). After attending a couple of them, I found out about the Summer of HPC and said if I have time in the coming summer, I will definitely apply! It will give me the chance to dig in to practical programming experience. In the end, I applied for the Performance of Parallel Python Programs on ARCHER2 … and I got IN!
So now, we (see us in the group photo – all motivated in the HPC home experience workshop) are doing a preparation week of learning how to use HPC, how to combine it with Python, C and Fortran, and how to run parallel programs with different tools such as Open Multi-Processing (OpenMP).
Can’t wait now for the next step of starting the project at the EPCC in Edinburgh (get some details here)!!!! WAIT there was something special about this year, COVID-19,… Well, great, I will do it from home and enjoy the flexibility. And the project is still the same. It is about performing improvement of a parallel-running Python version of a Computational Fluid Dynamics simulation of fluid flow in a cavity. Although at first we will just focus on the speed and not the outcome, I will try to bring that into my research after we managed to improve the code.
My goals for the Summer of HPC are:
- do some good team work with my project partner Jiahua
- enjoy the help of EPCC-team
- improve speed and options of the python code
- learn a new language like C (is already on my TODO list for a long time now)
- use HPC for research
Goals are set, we are ready to run.
Talk to you soon – here in the BLOG!
–> leave some comments on what do you want to hear about & what your interests are. Happy to answer your questions!