PI-HPC-Bench is a collection of High Performance Computing (HPC) benchmarks instrumented with Paramount Iterations (PI). It integrates widely used benchmark suites and real-world applications to evaluate HPC systems.
- NAS Parallel Benchmarks (NPB)
- Exascale Proxy Applications (ECP Proxy Apps)
- LLNL ASC Proxy Apps
- Real Applications: LAMMPS, ExaML
Our Paramount Iterations (PI) library source code is located in the /utils/ directory. Each benchmark directory includes an execution script at /exec_scripts/run_exec.sh, which is used to run the benchmark applications.
@article{https://doi.org/10.1002/cpe.7540,
author = {Aparecida Silva Camacho, Thais and Martins do Rosario, Vanderson and Oliveira Napoli, Otávio and Borin, Edson},
title = {PB3Opt: Profile-based biased Bayesian optimization to select computing clusters on the cloud},
journal = {Concurrency and Computation: Practice and Experience},
volume = {35},
number = {18},
pages = {e7540},
keywords = {Bayesian optimization, cloud computing, high-performance computing},
doi = {https://doi.org/10.1002/cpe.7540},
url = {https://onlinelibrary.wiley.com/doi/abs/10.1002/cpe.7540},
eprint = {https://onlinelibrary.wiley.com/doi/pdf/10.1002/cpe.7540},
year = {2023}
}
This project was developed at the Institute of Computing - Unicamp as part of @thaisacs master dissertation. You are free to use this code under the MIT LICENSE.