HPC, Big Data and future technologies for AI Applications in Agriculture

Updated: Aug 3, 2021

The EU-funded research project CYBELE converges the technology of High Performance Computing (HPC) and Cloud Computing to be applied in the domain of agriculture. Artificial Intelligence (AI) technologies have been utilized for crop yield prediction, disease detection, etc, to better understand the growth of crops and improve efficiency of production [1].

CYBELE proposed a hybrid architecture that enables the synergy of HPC and cloud systems as given in the figure. The Cloud clusters host long-running services that provide graphic interfaces for designing AI application pipelines and workflows. The applications with intensive computation are scheduled to execute on HPC clusters where program performance can be significantly enhanced. More information is given in [1,2].

CYBELE hybrid architecture is composed of big data and HPC clusters.

(photo copyright of HAWK HPC cluster: Ben Derzian for HLRS)

In the future, more advanced technologies will be introduced, such as Quantum Computing. The research that HLRS carries out on Quantum computing offers a great potential to advance computation on current HPC infrastructures to a new level. This technology together with AI may bring agriculture to a new dimension in the imminent future.

[1] Naweiluo Zhou, Li Zhong, Dennis Hoppe, "CYBELE: A Hybrid Architecture for HPC and Big Data for AI applications in Agriculture." HPC, Big Data, AI Convergence Toward Exascale: Challenge and Vision. Edited by Olivier Terzo, Jan Martinovic, CRC Press, 2021

[2] Zhou, N., Georgiou, Y., Pospieszny, M. et al. Container orchestration on HPC systems through Kubernetes. J Cloud Comp10, 16 (2021).