This repository collects the raw data, the scripts, and the plots we used for the following paper:
Andre Merzky, Mikhail Titov, Matteo Turilli, Ozgur Kilic, Tianle Wang, Shantenu Jha, “Scalable Runtime Architecture for Data-driven, Hybrid HPC and ML Workflow Applications." In Proceedings of HPC for AI Foundation Models & LLMs for Science (HPAI4S'25), co-located with The 39th IEEE International Parallel & Distributed Processing Symposium (IPDPS), June 3-7, 2025.
This paper concerns some of the use cases of the Low-Dose Understanding, Cellular Insights, and Molecular Discoveries (LUCID) project, and the experiments that characterized the LUCID infrastructure's performance.