P. Chris Broekema

P. Chris Broekema, Rob V. van Nieuwpoort

Currently, data-intensive scientific applications require vast amounts of compute resources to deliver world-leading science. The climate emergency has made it clear that unlimited use of resources (e.g., energy) for scientific discovery is no longer acceptable. Future computing hardware promises to be much more energy efficient, but without better optimized software this cannot reach its full potential. In this vision paper, we propose a generic AI-driven co-design methodology, using specialized Large Language Models (like ChatGPT), to effectively generate efficient code for emerging computing hardware. We describe how we will validate our methodology with two radio astronomy applications, with sustainability as the key performance indicator. This paper is a modified version of our accepted SuperCode project proposal. We present it here in this form to introduce the vision behind this project and to disseminate the work in the spirit of Open Science and transparency. An additional aim is to collect feedback, invite potential collaboration partners and use-cases to join the project.

Héctor Cadavid, Vasilios Andrikopoulos, Paris Avgeriou et al.

Context: The challenges posed by the architecting of System of Systems (SoS) has motivated a significant number of research efforts in the area. However, literature is lacking when it comes to the interplay between the disciplines involved in the architecting process, a key factor in addressing these challenges.Objective: This paper aims to contribute to this line of research by confirming and extending previously characterized architecting harmonization practices from Systems and Software Engineering, adopted in an ultra-large-scale SoS. Method: We conducted a confirmatory case study on the Square-Kilometre Array (SKA) project to evaluate and extend the findings of our exploratory case on the LOFAR/LOFAR2.0 radio-telescope projects. In doing so, a pre-study was conducted to map the findings of the previous study with respect to the SKA context. A survey was then designed, through which the views of 46 SKA engineers were collected and analyzed. Results: The study confirmed in various degrees the four practices identified in the exploratory case, and provided further insights about them, namely: (1) the friction between disciplines caused by long-term system requirements, and how they can be ameliorated through intermediate, short-term requirements; (2) the way design choices with a cross-cutting impact on multiple agile teams have an indirect impact on the system architecture; (3) how these design choices are often caused by the criteria that guided early system decomposition; (4) the seemingly recurrent issue with the lack of details about the dynamic elements of the interfaces; and (5) the use of machine-readable interface specifications for aligning hardware/software development processes.

Héctor Cadavid, Vasilios Andrikopoulos, Paris Avgeriou et al.

The problems caused by the gap between system- and software-level architecting practices, especially in the context of Systems of Systems where the two disciplines inexorably meet, is a well known issue with a disappointingly low amount of works in the literature dedicated to it. At the same time, organizations working on Systems of Systems have been developing solutions for closing this gap for many years now. This work aims to extract such knowledge from practitioners by studying the case of a large-scale scientific instrument, a geographically distributed radio telescope to be more specific, developed as a sequence of projects during the last two decades. As the means for collecting data for this study we combine online interviews with a virtual focus group of practitioners from the organization responsible for building the instrument. Through this process, we identify persisting problems and the best practices that have been developed to deal with them, together with the perceived benefits and drawbacks of applying the latter in practice. Some of our major findings include the need to avoid over-reliance on the flexibility of software to compensate for incomplete requirements, hidden assumptions, as well as late involvement of system architecting, and to facilitate the cooperation between the involved disciplines through dedicated architecting roles and the adoption of unifying practices and standards.