Design, Testing, and Reliability of Memristive-based Compute-in-Memory Architectures

Organizers: Surendra Hemaram and Anteneh Gebregiorgis

Abstract: This special session explores emerging compute paradigms with a focus on memristive compute-in-memory architectures. It addresses key challenges in fault tolerance, device variability, hardware-software co-design, and post-manufacturing testing for reliable brain-inspired and AI computing systems.

Talks:

• Emerging Compute Paradigms – Opportunities and Challenges — Norbert Wehn

Papers:

• ECC-Assisted Fault Tolerance in Memristive Crossbar-based CIM Architectures: An Overview — Surendra Hemaram, Said Hamdioui and Anteneh Gebregiorgis
• Variability Aware Neuro-Memristive Networks: Hardware Software Co-design Approaches — Aleena Kabeer, Darshitha Dinesh and Alex James
• Testing RRAM-Based Brain-Inspired Architectures after Manufacturing: Challenges and Solutions — Leticia Maria Bolzani Poehls and Andre Lucas Chinazzo

Emerging & Reliable Computing for Scalable Intelligence

Organizer: Juergen Becker

Abstract: This session highlights emerging computing technologies and architectures for scalable and efficient intelligence. Topics span chiplet-based systems, service-oriented computing, efficient arithmetic for AI acceleration, and neural network inference on resource-constrained edge platforms.

Talks:

• Emerging Technologies & The Chiplet Revolution — Guohao Dai
• SoC – Service on a Chip — Markus Abel

Papers:

• Evaluating Block Arithmetic Formats for Efficient AI Inference Hardware — Stefanos Christopoulos, Paul Genssler, Hussam Amrouch and Christian Sauer
• Efficient Neural Network Inference on CPU-Centered Edge Device — Grace-Li Zhang

Sustainable Intelligence at the Edge: IoT, Embedded, and Cyber-Physical Systems for a Greener World

Organizers: Cathal Hoare and Tiziana Margaria

Abstract: This special session examines how embedded intelligence, IoT, and cyber-physical systems can enable more sustainable and resource-efficient computing. Contributions cover edge AI, industrial automation, federated systems, neuromorphic computing, efficient LLM inference, and intelligent decision-support applications.

Talks: —

Papers:

• TinySimpleNet: A SimpleNet for Unsupervised Industrial Anomaly Detection on Edge Devices — Luigi Capogrosso, Enrico Fraccaroli, Michele Magno and Franco Fummi
• On the Deployment of Egocentric Pose Estimation Algorithms at the Edge — Ferdinando Pompanin, Enrico Martini, Franco Fummi and Nicola Bombieri
• Sustainable SoftPLCs at the Edge: An Open-Source Path to Right-Sizing and Retrofitting Industrial Automation — Gioacchino Ivan Noto and Nicola Dall'Ora
• LLM-Driven Optimization of Large Language Model Inference — Amir Aminifar
• A Federated Cyber-Physical Edge Architecture for Smart Hospital Room Automation — Matteo Iervasi, Tawite Noe Marthe, Mascuud Mohamed Ali, Abdulaziz Abdulkadir Hassan and Florenc Demrozi
• Toward a Unified Memristive Edge Platform for Sensing, Security, and Neuromorphic Computing — Heba Abunahla
• Embedded Intelligence at the Edge: Recommender Systems, Efficient Simulation, and Operator-Centred Decision Support — Cathal Hoare and Tiziana Margaria

Edge AI for Smart Agriculture

Organizer: Theocharis Theocharides

Abstract: This session explores how Edge AI, embedded computing, and advanced sensing technologies can transform smart and precision agriculture. Topics include robotic and neuromorphic vision, low-power AI systems, precision monitoring, and on-device perception for efficient and sustainable agricultural applications.

Talks: —

Papers:

• Embedded Systems for Precision Agriculture Monitoring: An Electronics Designer's Perspective — Lorenzo Peppi, Alessandro Torrisi, Michele Gullino, Luigi Manfrini, Luisa Petti and Luca De Marchi
• Neuromorphic Vision Technologies in Smart Agriculture: Challenges and Opportunities — Luca Peres and Davide Bertozzi
• Edge AI for Smart and Precision Agriculture: An Overview of Robotic Vision for Field Perception and Decision Support — Theocharis Theocharides, Antonis Savva and Yiota Victoria Phakoukaki
• Pollin8: Species-Level Pollinator and Pest Counting on a Milliwatt System-on-Chip — Denisa-Andreea Constantinescu, Philip Wiese, Mattia Consani, Victor Kartsch, Luca Benini and David Atienza

DfT, Reliability and Functional Safety for Edge AI SoCs

Organizers: Maksim Jenihhin and Matteo Sonza Reorda

Abstract: This special session addresses the growing need for dependable and functionally safe Edge AI systems. It brings together advances in design-for-test, runtime fault detection and correction, reliability assessment, and safety-oriented processor testing and hardening across the computing stack.

Talks:

• Towards Dependable Edge AI: Reliability Assessment and Hardening Across the Compute Stack — Matteo Sonza Reorda

Papers:

• DfT-Assisted Testing of Resistive Short Defects in SRAM-Based In-Memory Computing Architectures — Dimitrios Stylianos Lampridis, Yiorgos Tsiatouhas, Maria K. Michael and Theocharis Theocharides
• NURSE: Syndrome-Guided Runtime Correction of Pipeline Interconnect Faults in RISC-V Processors — Ashwin Santhosh, Artur Jutman, Endri Kaja, Wolfgang Ecker and Maksim Jenihhin
• Towards ISO 26262-compliant Vector Processors: Combining STLs and LBIST — Gustavo Vilar de Farias, Dwijesh Kurada, Sergiu-Mohamed Abed, Juan-David Guerrero-Balaguera, Josie Esteban Rodriguez Condia, Ahmet Cagri Bagbaba, Felipe Augusto da Silva and Matteo Sonza Reorda