The 16th PPAM 2026 conference (Parallel Processing & Applied Mathematics) in Poznań, Poland, took place from August 30 to September 2, organized by the Department of Computer Science of the Częstochowa University of Technology. The conference featured workshops, tutorials, and talks around parallel and distributed processing as well as a Best Paper Award. We are delighted to announce that our colleagues Ayesha Afzal, Markus Manfred Li, and Michael Panzlaff were able to achieve this success for the main track.
The Award-Winning Paper: Targeting GPU Energy Efficiency
Adjusting the clock frequency or the power limit of GPUs changes both the processing speed and the energy required to solve a problem. The paper “Modeling and Chasing the Energy-Efficiency Sweet Spots in Modern GPUs” examines this behavior in more detail on current hardware (Nvidia A40, A100, H100, H200). Using HPC MD (molecular dynamics) applications and the stress case FIRESTARTER benchmark, the study demonstrates that there is a clear range for energy-optimized operation and that this range can be accurately modeled. In this case, optimization of the clock frequency is the preferred choice over power limit optimization. Find the full paper here: DOI: 10.48550/arXiv.2607.00819.
The award is granted by the PPAM Chairs and the Program Committee with a focus on originality and the importance of the research in the field of high-performance computing. A round of applause to Ayesha, Markus, and Michael for their successful work!
