ERP Data Analysis and Visualization in High-Performance Computing Environment

In the era of the fourth industrial revolution the enterprise resource planning system (ERP) becomes a foundation for interconnection between logistics systems, production facilities, smart machines, IoT-enabled devices and other enterprise data sources. The paper proposes an approach to extend the ERP integrated analytical tools capabilities by processing ERP data in a multi-tenant GPU-enabled high-performance computing (HPC) environment. Corporate analytic features in conjunction with GPU in-memory processing of big structured and unstructured data increase the performance and analysis effectiveness for enterprise machine learning (ML) tasks. The approach proposes sharing the data in GPU memory using open analytic platform along with existed ERP analytical capabilities on example of SAP S/4Hana. Considered solution accelerates data scientists work with ERP data sets and could be used for faster quality AI model creation and easier data interaction in unspecific for ERP visualization way like immersive learning with virtual or augmented reality (VR/AR).