Ferran Llorà Nieto
Bachelor's degree student in Computer Engineering
Characterization of CPU-GPU workload performance in heterogeneous high-performance computing systems
Characterization of CPU-GPU workload performance in heterogeneous high-performance computing systems
Description
This project focuses on the characterization and performance analysis of collaborative CPU-GPU applications on large-scale high-performance computing (HPC) systems. The goal is to understand how workloads behave on heterogeneous platforms, identifying bottlenecks and optimization opportunities. By evaluating real applications and benchmarks, Ferran contributes to improving the efficiency of HPC systems that combine CPUs and GPUs, providing essential knowledge for the co-design of future hardware and software infrastructures for computational biology and genomics.Background
Ferran is currently in the final year of his Bachelor's degree in Computer Engineering, where he has specialized in high-performance computing, parallel programming and computer architecture. In parallel to his academic training, he is acquiring practical research experience at the Barcelona Supercomputing Center, where he applies his knowledge to the study of heterogeneous HPC systems. This combination of formal training and applied research prepares him to contribute to optimization and performance characterization tasks in collaborative CPU-GPU workloads.Motivation
Ferran chose this research because he enjoys solving challenging problems and learning new things. The project offers him the opportunity to work at the intersection of advanced computing and real-world applications, while gaining valuable experience that will undoubtedly be useful in his future career. His motivation is to better understand how heterogeneous HPC systems can be exploited to accelerate demanding scientific workloads. Ferran's research interests include high-performance computing, collaborative CPU-GPU applications, performance characterization and optimization of heterogeneous systems.
Host Organization
Supervisors
Santiago Marco-Sola
UPC supervisor
The content of this website reflects only the views of the Catedra Chip Chair UPC project.
