# Academic Salon on Low-Latency Communication, Programmable Network Components and In-Network Computation ## Aims * Core technologies on Low-Latency Communication, Programmable Network Components, and use-cases for In-Networked Computation * Applications, in particular from the industrial domain, with requirements met by these core technologies, and that benefit from the capability of in-network computation. * Network-related innovations, contributions on implementation, innovative memory and processing technologies, methods and tools for performance assessment, and insights into industrial and automotive applications. ## Topical Scope * Network-related innovations, in particular architecture innovations such as acceleration techniques for programmable Network Interface Cards and Switches * Innovative processing and memory technologies, such as P4-based protocol processing * Design, specification, verification, implementation, measurement, testing, and analysis of programmable network components * Methods and tools for performance assessment * Applications and use-cases, in particular of industrial and the automotive domain. ## Organisation by Technical University of Munich * Georg Carle (Chair) * Andreas Herkersdorf, Jörg Ott (Vice-Chairs) ## Venue * Afternoon 29 and Morning 30 September 2022, Online event (Virtual Munich) ## Tools * Presentations and Panel Discussions are organized in Zoom. * Zoom instructions: https://codimd.net.in.tum.de/s/Hy2TudmHt * Zoom dialin: * Webinar-ID 676 2935 9907, passcode 663900, Link: https://tum-conf.zoom.us/j/67629359907 * Telephone Dialin +49 69 3807 9884, Webinar-ID: 676 2935 9907 * Link for presenters to upload slides: https://oc.net.in.tum.de/index.php/s/9jH32e2SaqEcMrZ * Slides will be available after the event here: * https://oc.net.in.tum.de/index.php/s/BfD2PGsgxCNMX3N # Program ## Thursday 29.9.2022 ### Session 1 - 14:00 - 15:45 #### Topic Area: Network Paradigms that benefit from Programmable Network Components ##### Session Chair: Georg Carle * Marco Canini, KAUST: Scaling Deep Learning and Datacenter Applications with Programmable Networks * Steffen Lindner, Daniel Merling, Michael Menth, Universität Tübingen: P4-Based Implementation of BIER and BIER-FRR for Efficient Multicast * Luigi Iannone, Huawei Research Paris: Topological Addressing ### Session 2 - 16:00 - 17:30 #### Topic Area: Real-Time Networking approaches and tools for Industrial Application Scenarios ##### Session Chair: Georg Carle * Paolo Medagliani, Huawei Research Paris: High capacity and Resilient Large-scale Deterministic IP Networks * Filip Rezabek, Marcin Bosk, TUM: Reproducible Layer3-enabled TSN Experiments * Panel Discussion: What challenges of industrial networking remain to be resolved? https://hedgedoc.net.in.tum.de/GPpoQ32QThysrl6vtt2QuA ## Friday 30.9.2022 ### Session 3 - 9:00 - 11:00 #### Topic Area: Programmable Network Interface Cards and In-Network Computation - In-network computation in industrial scenarios, Processor-based Smart-NICs ##### Session Chair: Michael Menth * Dejan Kostic, KTH Stockholm: Packet Ordering for Improving Application Performance using SmartNICs * Francesc Fons, Angela Gonzalez Mariño, Abdoul Aziz Kane, Huawei Research Munich: The paradigm shift in automotive zonal gatewaying * René Glebke, Klaus Wehrle, RWTH Aachen: In-network computation and Processor-based Smart-NICs ### Session 4 - 11:15 - 13:00 #### Topic Area: Mechanisms and Methods for Real-Time Networking approaches suitable for Industrial Application Scenarios ##### Session Chair: Jörg Ott * Lisa Maile, University Erlangen-Nürnberg: QoS mechanisms for Industrial Applications * Max Helm, Florian Wiedner, TUM: Tail Latency Estimation and Verification * Mubarak Adetunji Ojewale, University of Porto: Multi-Level Preemption in TSN * Panel Discussion Structure: Programmable network components, tool chains, and applications https://hedgedoc.net.in.tum.de/jbqbgHxiTQyLzU5dzBUBgw * Panel Discussion Structure: What is the future of In-Network Computation? https://hedgedoc.net.in.tum.de/4gkNCa8TS4KiH8ZCES6r0A * Wrapup of the Academic Salon