Paper M100 ExaData: a data collection campaign on the CINECA’s Marconi100 Tier-0 supercomputer

Exciting news from our DECICE project!

Our colleagues at the University of Bologna have published a groundbreaking scientific paper in Nature Scientific Data, unveiling the “M100 ExaData: a data collection campaign on the CINECA’s Marconi100 Tier-0 supercomputer.” This paper marks the creation of the first-ever open dataset for HPC telemetry, a significant milestone for the advancement of AI digital twins in HPC and cloud environments.

Over a ten-year-long project, the University of Bologna research team at the DEI and DISI department designed the monitoring framework (EXAMON) deployed at the Italian supercomputers in CINECA’s datacenter. This dataset unveils a holistic view of the tier-0 Top10 Marconi100 supercomputer, encompassing management, workload, facility, and infrastructure data gathered over two and a half years of operation. The dataset, available via Zenodo, is an impressive achievement, being the largest ever made public, with a size of 49.9TB before compression. To simplify data access, they also provide open-source software modules, along with direct usage examples.

Read the paper and explore the potential of our groundbreaking dataset: https://www.nature.com/articles/s41597-023-02174-3

We’re excited to contribute to the development of AI digital twins and propel the field of HPC telemetry. Stay tuned for more remarkable advancements from our DECICE project!

Author: Andrea Bartolini

Links

Paper: https://www.nature.com/articles/s41597-023-02174-3

University of Bologna: https://www.unibo.it/en

Keywords

DECICEproject | AI | HPC | DigitalTwins | NatureScientificData