Application-Driven Exascale: The JUPITER Benchmark Suite
Abstract
Benchmarks are essential in the design of modern HPC installations, as they define key aspects of system components. Beyond synthetic workloads, it is crucial to include real applications that represent user requirements into benchmark suites, to guarantee high usability and widespread adoption of a new system. Given the significant investments in leadership-class supercomputers of the exascale era, this is even more important and necessitates alignment with a vision of Open Science and reproducibility. In this work, we present the JUPITER Benchmark Suite, which incorporates 16 applications from various domains. It was designed for and used in the procurement of JUPITER, the first European exascale supercomputer. We identify requirements and challenges and outline the project and software infrastructure setup. We provide descriptions and scalability studies of selected applications and a set of key takeaways. The JUPITER Benchmark Suite is released as open source software with this work at https://github.com/FZJ-JSC/jubench.
Cite
@article{arxiv.2408.17211,
title = {Application-Driven Exascale: The JUPITER Benchmark Suite},
author = {Andreas Herten and Sebastian Achilles and Damian Alvarez and Jayesh Badwaik and Eric Behle and Mathis Bode and Thomas Breuer and Daniel Caviedes-Voullième and Mehdi Cherti and Adel Dabah and Salem El Sayed and Wolfgang Frings and Ana Gonzalez-Nicolas and Eric B. Gregory and Kaveh Haghighi Mood and Thorsten Hater and Jenia Jitsev and Chelsea Maria John and Jan H. Meinke and Catrin I. Meyer and Pavel Mezentsev and Jan-Oliver Mirus and Stepan Nassyr and Carolin Penke and Manoel Römmer and Ujjwal Sinha and Benedikt von St. Vieth and Olaf Stein and Estela Suarez and Dennis Willsch and Ilya Zhukov},
journal= {arXiv preprint arXiv:2408.17211},
year = {2024}
}
Comments
To be published in Proceedings of The International Conference for High Performance Computing Networking, Storage, and Analysis (SC '24) (2024)