Skip to content

Bine, PhD student

My name is Bine Brank and I am a 24 year old PhD student at the Jülich Supercomputing Centre. I originally come from Ljubljana, Slovenia, and I moved to Germany in 2016 for my master’s degree at the University of Wuppertal. In my free time I like to practice martial arts, bouldering or hiking in nature. I am also interested in mobile game development, aviation and artificial intelligence.

Jülich

Although my background is primarily in physics, I have a long-time interest in high performance computing. After I finished my master’s degree in computer simulation, I knew that I wanted to gain more knowledge about the state-of-the-art computers needed by scientists to run their computationally expensive applications. Since Jülich is one of Europe’s leading supercomputing centres, the decision to choose it for my PhD was not hard to make. Soon after starting my PhD, I joined the Mont-Blanc 2020 project as a part of my research on computer architectures. It came as a surprise rather than something I had planned, and I have to admit that it seemed quite a burden at the beginning, since I had almost zero prior experience. But  I soon started to understand my role and appreciate that I was a part of something big.

I soon started to understand my role and appreciate that I was a part of something big

Jülich acts as a co-design partner in the Mont-Blanc 2020 project. Because the institution has a long-term expertise in providing supercomputing services to the HPC community, it has an in-depth knowledge of user demands for future supercomputers. A broad range of available scientific applications are used for modelling and simulation of system performance. Additionally, Jülich also offers a wide range of different machines and thus provides excellent conditions for computer architecture research. Jülich’s involvement in Mont-Blanc 2020 is beneficial to gain advance knowledge on the future solutions that will be used in their systems.

My work

My work primarily focuses on Arm’s Scalable Vector Extension instruction set architecture. A major part of my research involves the usability of SVE for a selected set of relevant HPC applications. This involves porting applications to the new architecture and then testing them in simulators such as gem5 for computer architecture research. Besides, I am also developing tools that will facilitate the development of SVE code. This includes a code generator that will allow a flexible generation of assembler code, and a low level SIMD library that will encapsulate the complex SVE kernels. This work requires a thorough understanding of the Arm SVE instruction set architecture and low-level programming. It gives me an excellent chance to contribute to cutting-edge research and learn about new things in this fast-developing field.

Participating in Mont-Blanc 2020 is also an opportunity for me to see first-hand how big European-funded projects are organized and managed. This is not something you learn at university, although it plays a very important part for today’s research community. This is a very valuable experience for any individual who is at the beginning of his career. Furthermore, a large multicultural consortium of high-tech European companies such as Mont-Blanc 2020 gathers a lot of expertise and provides an excellent insight into industry and how new technologies are being developed. Not every PhD student has a chance to be so directly involved in such projects and I am thankful for this opportunity.

This is not something you learn at university

I feel that the decision to participate in this project was definitely the right one. I can already say that I have gained a lot of knowledge and much needed experience since joining the project one year ago, and I am looking forward to the challenges ahead.