On April 23^rd and 24^th was held a brainstorming architecture meeting focussing on two technical pillars/functionalities of the AMAZING-6G project: compute continuum and dynamic network slicing.

The first full day was involving main actors in those two fields: UoS as WP2 lead, IMEC, Cap Gemini, and TNO.

We started from the preliminary architecture design and elaborated on it, seeking consistent end-to-end workflow from start (consisting of Vertical intents) to finish (deployment of Vertical services wherever optimally executed, including near a UPF).

This work was later on consolidated by elaborating different sequence diagrams, elucidating how the different functional components of the architecture would be collaborating with each other (and with the Vertical) in order to deliver the two aforementioned major functionalities.

Indeed, dynamic network slicing allows a Vertical 1/ to express in informal form its QoS networking requirements and 2/ to get eventually allocated automatically a customised network slice meeting its targeted KPIs.

As far as compute is concerned, expressing the compute requirement of its own services via intents, leads to an optimal “Task—> Node” allocation distribution of the various services among available compute nodes.

The allocation itself considers a wide range of parameters such as the nature and ownership of available compute nodes, their power sources and overall compute characteristics, plus the Vertical services’ compute requirements. Complementary post-deployment close loop monitoring ensures continuity of service execution and its expected performance levels.

Second day half-day of the workshop focussed on internal project-wise dissemination and technical follow-up discussions.