Federated Edge–HPC architectures for AI workflows in privacy-sensitive and real-time domains

Spyridopoulos, Theodoros ORCID: https://orcid.org/0000-0001-7575-9909, Rana, Omer ORCID: https://orcid.org/0000-0003-3597-2646, Majib, Yasar, Belarbi, Othmane ORCID: https://orcid.org/0000-0002-6106-7669, Szydlo, Tomasz, Nandan Jha, Devki, Li, Yinhao, Ullah, Rehmat, Kumar, Amrit and Ranjan, Rajiv 2026. Federated Edge–HPC architectures for AI workflows in privacy-sensitive and real-time domains. [Project Report]. National Federated Compute Services. Available at: https://nfcs-networkplus.ac.uk/projects/federated-...

PDF - Accepted Post-Print Version Download (1MB)

Abstract

The project explores the feasibility of linking national HPC resources, such as Isambard AI, with edge infrastructures at Cardiff and Newcastle Universities. Its central aim is to build a community of stakeholders through workshops that identify application requirements, shape use case narratives, and guide a future roadmap for HPC–Edge federation. Complementing this, the project will trial and assess workflows for distributing AI workloads across heterogeneous environments, with a focus on data privacy, decentralised orchestration, and real-time decision support. A testbed exercise will further examine the opportunities and challenges of such a federation.

Item Type:	Report (Project Report)
Status:	Published
Schools:	Schools > Computational & Mathematical Sciences
Publisher:	National Federated Compute Services
Date of First Compliant Deposit:	17 August 2026
Last Modified:	17 Aug 2026 09:00
URI:	https://orca.cardiff.ac.uk/id/eprint/189006

Actions (repository staff only)

Edit Item