Skip to content

Spotlight on Young Researchers: Using supercomputers with mathematical modelling to solve science and engineering problems

 

In the last decades, “classic” computers have played a vital role in science. As technology advances, access to quantum computers is rising, but not without challenges, as using them is not easy. One aspect researchers are working on is using supercomputers to run numerical simulations, which can be used to help solve a variety of science and engineering problems.

In the last decades, “classic” computers have played a vital role in science. As technology advances, access to quantum computers is rising, but not without challenges, as using them is not easy. One aspect researchers are working on is using supercomputers to run numerical simulations, which can be used to help solve a variety of science and engineering problems.

For example, we do have mathematical modelling for weather modelling and supercomputers. Still, if we do not know how to execute that mathematical modelling on the supercomputers, we cannot predict the weather in the coming days. Imagine if you run that mathematical modelling using a normal laptop: we cannot predict the weather for the next few days. Until now, I have worked with how to run the simulations for fluid dynamic problems (for example, aeroplane simulation and flood modelling), tsunami simulation and plasma physics. If we do not run those simulations faster, innovative mathematical modelling and powerful supercomputers are not used.
Ezhilmathi Krishnasamy Postdoc and computer scientist working at the University of Luxembourg.

Ezhilmathi combines mathematical modelling with high-performance computing. He explains that science and engineering problems are defined by partial differential equations, and those equations are approximated by numerical methods.

“Finally, these numerical methods need to be solved efficiently and faster. This is where I play a crucial role as a researcher because, using the supercomputer, these numerical methods need to be solved.”

“It might look simple, but it is a huge domain, and many things must be considered. The core research areas are parallel programming, heterogeneous computer architecture and numerical methods. As researchers or end users, we need to find an optimal solution for using supercomputers for numerical methods.”

Mostly working on how to run numerical simulations efficiently on supercomputers – where it is complex and heterogeneous with CPU, GPU and other accelerators – Ezhilmathi’s work address a variety of research questions.

“In the future, I would like to focus on plasma physics and quantum computing. Both of these research directions are not that easy and are in the development stage. Therefore, lots of efforts are needed in the future to explore more about plasma physics and quantum computers.”

Ezhilmathi Krishnasamy is a Postdoc and computer scientist working on the FNR supported project HEXAPIC. He is based in the Parallel and Optimization Group at the University of Luxembourg, headed by Pascal Bouvry.

MORE ABOUT EZHILMATHI KRISHNASAMY

How he describes himself as a researcher

“I feel I have some purpose in my life or some responsibility. For example, the things that I enjoy in my life currently have been given to me by someone who dedicated their life, such as Airplanes, Medical advancements, Computers, and electronics. Therefore, I want to try my best to contribute as much as possible to science. I have been working with scientific computing, using supercomputers for mathematical modelling. One specific direction would be Plasma Physics, where I intend to do the computer simulation. Because using experimental Plasma Physics simulation would be really hard and not even feasible. My team focuses on the HPC; I initially got the job offer after my PhD, and I am in Luxembourg.”

What he loves about science

“Without science, we would not have anything that we have now, for example, medicine, engineering and computers. So, science is one of the fundamentals for humans to survive. I am happy that I can also contribute to science, which could benefit the present or future generations.”

Mentors with an impact

“I have encountered many people during my studies, and during that time, I have encountered many professors and mentors. All of them, in a specific way, motivated me to stay in academics and encouraged me to do research and teach.”

Where he sees himself in 5 years

“I would like to stay in academia, doing research and teaching. I know it is hard to get an academic position; however, I will do my best to find a good one in Luxembourg or elsewhere.”

Related contacts

Emily Iversen

Digital Communication Manager