Providing services is one of the central ideas of the NFDI. The Marketplace of PUNCH4NFDI will serve as a visible interface to the entire NFDI. As a result of the cooperative work in the different TAs, and based on the deliverables PUNCH4NFDI will develop within the upcoming 5 years, the consortium will provide several large-scale distributed services for the processing and management of scientific data to the users.

The PUNCH4NFDI services have been grouped into the following 5 service classes:

These service classes are described below. Details about these service classes and individual services can be found by clicking the titles.

Service Class 1: PUNCH Central Infrastructure

Central to PUNCH4NFDI is comfortable access for the users to all the provided services, offered via the PUNCH Science Data Platform (PUNCH-SDP). All other services mentioned below are an integral part of this single, dedicated platform. Through the marketplace and the data portal these services will be published, including the AAI prototype.

Service Class 2: Data access and management services

Access to PUNCH4NFDI open data archives such as the open data of CERN’s LHC experiments and the IVOA, but also smaller data collections will be offered in the data  lake. Dynamic disk cache technology for the integration of opportunistic storage  resources will be provided. Data processing will be supported by FTS and Rucio services  for evaluation purposes. Specifically, corresponding reference guides will support data producers in publishing their data and software.

Service Class 3: Analysis software and data irreversibility services

For the extraction of high-level information, tools can be found in the PUNCH software repository. Among others, this contains data analysis and simulation routines optimised  for multi-GPU systems, a framework for AutoML on scientific data, and a framework for  conversion/reading of data for combined analyses on heterogeneous systems. For real- time applications, algorithms optimised for sorting, hardware-specific clustering and  pattern recognition, and methods for transforming dynamical archive queries into  dynamic filters (and vice versa) will be developed, as well as a generalised toolkit for  predictive maintenance and anomaly detection.

Service Class 4: Metadata services

The possibility to build from the research products of PUNCH4NFDI a dynamic knowledge  fabric is the next step. The definitions of their functionality and interfaces, prototype  metadata schemes, data formats and published and interoperable digital research  products using the full range of PUNCH4NFDI services are provided. The corresponding  catalogue technology will be accessible for all interested users through the PUNCH-SDP.

Service Class 5: Computing and storage resource services

For efficient data management and analysis, IT resources as provided by  Compute4PUNCH are needed. Therefore, interfaces to existing infrastructures are  provided, for example, to the supercomputer HLRN. Additionally, small fractions of  community-specific resources jointly managed by the COBalD/TARDIS compute resource  management software framework will be made available to PUNCH4NFDI and beyond for  data analysis. Finally, prototyping interactive analysis via multi-cloud resources will be  facilitated for interested users.