Room: TEK 1.023

11h00 - 12h30: Session 1, Chair: Christian Perez

•   11h00-11h15h Welcome: Christian Perez

•   11h15 - 12h30 Invited talk: Kate Keahey - Building Scientific Instruments in the Edge to Cloud Continuum

12h30 - 14h00: Lunch

14h - 15h30: Session 2, Chair: Gabriel Antoniu

• 14h00 - 14h30 IoT Orchestration in the Compute Continuum: Integrating Kubernetes with Stack4Things - Luca D'Agati, Giuseppe Tricomi, Michele Arena, Francesco Longo, Antonio Puliafito and Giovanni Merlino

• 14h30 - 15h00 Towards a flexible Network Operating System Testbed for the Computing Continuum - Julien Caposiena, Oscar Carrillo, Frédéric Le Mouël, Baptiste Jonglez, Pierre Neyron and Thierry Arrabal

• 15h00 - 15h30 Towards NDN-based end-to-end resource reservation for Computing Continuum - Thierry Arrabal, Julien Caposiena, Frédéric Le Mouël, Stéphane Frénot and Oscar Carrillo

15h30 - 16h00: Break

16h00 - 17h30: Session 3, Chair: Kate Keahey

• 16h00 - 16h30 Does Quantization Improve Inference Speed? It Depends - Ahmed Farrukh, Mohamed Saeed and Fraida Fund

• 16h30 - 17h00 Experimenting IoT-edge-cloud-HPC continuum on existing platforms - Cassandre Vey, Adrien van den Bossche, Réjane Dalcé, Georges Da Costa, Olivier Negro and Thierry Val

• 17h00 - 17h30 Invited talk: Gabriel Antoniu - Challenges for building the Edge-Cloud-HPC Data Continuum: the InPEx perspective

The emergence of new opportunities in AI, networking, as well as edge computing and IoT creates the potential for new avenues of exploration in science but at the same time challenges our understanding of the role and shape of infrastructure. An effective exploration of the edge to cloud continuum requires research infrastructure where those challenges can be investigated, and where new solutions and disruptive ideas can be developed, deployed, tested, and shared. Such infrastructure has to not only support a diversity of hardware configurations, deployments at scale, as well as deep reconfigrability so that a wide range of experiments can be supported – but also mechanisms and services that will allow the community to share repeatable digital artifacts so that new experiments and results can be easily replicated and help enable further innovation.

This type of research infrastructure (RI) has been implemented in systems like Chameleon, SLICES, FABRIC, or the PAWR testbeds, many of which address distinct, if overlapping, aspects of this problem. This workshop will provide a forum for the RI designers, engineers, and users to discuss emergent use cases and modes of usage; RI solutions and and operations; features that support research in the computing continuum; user interfaces and usability; experimental support for reproducibility; building and sustaining RI user communities; as well as evaluating and sustaining RI.