A Proposal for a Simulation of PCAD Network Topology Using SimGrid
Resumo
Simulation is a well-established methodology for evaluating High-Performance Computing (HPC) systems, ensuring reproducibility and reduced energy footprints. This work proposes a programmatic network model of the PCAD infrastructure, a heterogeneous institutional cluster, using the SimGrid S4U API. Focusing on four core partitions, we map the physical architecture into a simulated structure and calibrate the platform properties with experimentation. The main goal is to enable local researchers to replicate experiments and study scalability even with limited physical hardware.
Referências
Casanova, H., Giersch, A., Legrand, A., Quinson, M., and Suter, F. (2025). Lowering entry barriers to developing custom simulators of distributed applications and platforms with SimGrid. Parallel Computing, 123:103–125.
Cornebize, T. (2021). High Performance Computing : towards better Performance Predictions and Experiments. Theses, Université Grenoble Alpes [2020-....].
Dongarra, J. J., Luszczek, P., and Petitet, A. (2003). The LINPACK benchmark: past, present and future. Concurrency and Computation: Practice and Experience, 15(9):803–820.
INF/UFRGS (2025). Laboratório de processamento paralelo e distribuído (lppd). [link]. Accessed: 2026-02-16.
Jain, R. (1991). The Art of Computer Systems Performance Analysis: Techniques For Experimental Design, Measurement, Simulation, and Modeling, NY: Wiley.
Karlin, I., Keasler, J., and Neely, J. R. (2013). Lulesh 2.0 updates and changes. Technical report, Lawrence Livermore National Laboratory (LLNL), Livermore, CA (United States).
Melo, A. A. F. (2025). Um simulador da rede do supercomputador frontier utilizando simgrid. Bachelor’s thesis.
Top500 (2022). Frontier. [link]. Accessed: 2026-02-22.
