TY  - JOUR
AU  - Ejarque, Jorge
AU  - Badia, Rosa M.
AU  - Albertin, Loïc
AU  - Aloisio, Giovanni
AU  - Baglione, Enrico
AU  - Becerra, Yolanda
AU  - Boschert, Stefan
AU  - Berlin, Julian R.
AU  - D’Anca, Alessandro
AU  - Elia, Donatello
AU  - Exertier, François
AU  - Fiore, Sandro
AU  - Flich, José
AU  - Folch, Arnau
AU  - Gibbons, Steven J.
AU  - Koldunov, Nikolay
AU  - Lordan, Francesc
AU  - Lorito, Stefano
AU  - Løvholt, Finn
AU  - Macías, Jorge
AU  - Marozzo, Fabrizio
AU  - Michelini, Alberto
AU  - Monterrubio-Velasco, Marisol
AU  - Pienkowska, Marta
AU  - de la Puente, Josep
AU  - Queralt, Anna
AU  - Quintana-Ortí, Enrique S.
AU  - Rodríguez, Juan E.
AU  - Romano, Fabrizio
AU  - Rossi, Riccardo
AU  - Rybicki, Jedrzej
AU  - Kupczyk, Miroslaw
AU  - Selva, Jacopo
AU  - Talia, Domenico
AU  - Tonini, Roberto
AU  - Trunfio, Paolo
AU  - Volpe, Manuela
TI  - Enabling dynamic and intelligent workflows for HPC, data analytics, and AI convergence
JO  - Future generation computer systems
VL  - 134
SN  - 0167-739X
CY  - Amsterdam [u.a.]
PB  - Elsevier Science
M1  - FZJ-2022-02034
SP  - 414-429
PY  - 2022
AB  - The evolution of High-Performance Computing (HPC) platforms enables the design and execution of progressively larger and more complex workflow applications in these systems. The complexity comes not only from the number of elements that compose the workflows but also from the type of computations they perform. While traditional HPC workflows target simulations and modelling of physical phenomena, current needs require in addition data analytics (DA) and artificial intelligence (AI) tasks. However, the development of these workflows is hampered by the lack of proper programming models and environments that support the integration of HPC, DA, and AI, as well as the lack of tools to easily deploy and execute the workflows in HPC systems. To progress in this direction, this paper presents use cases where complex workflows are required and investigates the main issues to be addressed for the HPC/DA/AI convergence. Based on this study, the paper identifies the challenges of a new workflow platform to manage complex workflows. Finally, it proposes a development approach for such a workflow platform addressing these challenges in two directions: first, by defining a software stack that provides the functionalities to manage these complex workflows; and second, by proposing the HPC Workflow as a Service (HPCWaaS) paradigm, which leverages the software stack to facilitate the reusability of complex workflows in federated HPC infrastructures. Proposals presented in this work are subject to study and development as part of the EuroHPC eFlows4HPC project.
LB  - PUB:(DE-HGF)16
UR  - <Go to ISI:>//WOS:000808123100004
DO  - DOI:10.1016/j.future.2022.04.014
UR  - https://juser.fz-juelich.de/record/907436
ER  -