Wednesday, November 26th
08:00 Registration
08:30 Session 1- Opening
09:00 Session 2 Keynote: Roger Espasa Roger Espasa (Semidynamics): Perspectives on the Semiconductor and Fabless industry
09:30
10:00
10:30 Coffee Break & Posters
11:00 Session 3A: Hardware and Software for RISC-V Session 3B: Sensing
19:30
12:00
12:30 Session 4A: Security and Power Systems Session 4B: Signal Processing and Power Systems
13:00
13:30
14:00 Lunch
14:00
14:00
15:30 Session 5A: AI Circuits and Systems Session 5B: Neuromorphic Circuits, Systems and Technologies I
16:00
16:30
17:00 Session 6A: Neuromorphic Circuits, Systems and Technologies II Session 6B: System-Level Analysis and Exploration
17:30
18:00
18:30 Free time
19:00 Welcome Reception
19:30
20:00
Thursday, November 27th
08:00 Registration
08:30
09:00 Session 8 Keynote: Alberto Bosio (Ecole Centrale de Lyon): Trustworthy AI: The role of the Hardware
09:30
10:00
10:30 Coffee Break & Posters
11:00 Session 9A: AI-Driven Development of High-Performance Electronic Systems and Applications-1 Session 9B: Monitoring and Control
19:30
12:00
12:30 Session 10A: Circuit Design and Analysis Session 10B: Powering circuits
13:00
13:30
14:00 Lunch
14:00
14:00
15:30 Session 11 Panel: Industry-University Collaboration and Technology Transfer
16:00
16:30
17:00
17:30 Free time
18:00 Social Event
18:30
19:00
19:30
20:00
20:30
21:00
21:30
Friday, November 28th
08:00 Registration
08:30
09:00 Session 13 Keynote: Ivan Rebollo (NXP): Dual Perspective on Non-Volatile Memory and RFID: Technological Progress and Engineering Challenges
09:30
10:00
10:30 Coffee Break & Posters
11:00 Session 14A: Hardware Accelerators Session 14B: RF & Communications: Sensing
11:30
12:00
12:30 Session 15A: AI-Driven Development of High-Performance Electronic Systems and Applications-2 Session 15B: Quantum and low power
13:00
13:30
14:00 Session 16: Closing Ceremony
14:30 Lunch
15:00
15:30

Session 1: Registration & opening
Chairs: Eugenio Villar and Frédéric Pétrot

Wednesday, November 26th
Session 2 Keynote 1: Roger Espasa (Semidynamics): Perspectives on the Semiconductor and Fabless industry
Chair: Eugenio Villar
Abstract: In this talk we’ll discuss the semiconductor industry focusing on its fabless component and analyze it from a European perspective. A brief discussion on the different players within the semiconductor industry will be followed by thoughts and remarks on the current status of the fabless and foundry ecosystem in Europe

Session 3A: Hardware and Software for RISC-V
Chairs: Francesc Moll and Borja Pérez
11:00 Sara Alonso, Alejandro Arteaga, Leire Muguira, Carlos Cuadrado, Aitzol Zuloaga, Jaime Jiménez, Jesús Lázaro, José Ignacio Gárate, José Angel Araujo, Victor Martínez, Unai Bidarte and Armando Astarloa
A proof-of-concept ASIC RISC-V based SoC for Industrial Applications (abstract)
11:30 David Cantero, Alex Ugena, Laura Sanz, Alejandro Arteaga and Armando Astarloa
Performance Analysis of Convolution Function for IA Edge Computing Acceleration using a 32-bit RISC-V CPU Implementation (abstract)
12:00 Roman Cardenas, Pedro Malagon, Patricia Arroba, Josue Pagan and Jose M. Moya
RISCV-SLIC: Rust Software Level Interrupt Controller for RISCV microcontrollers (abstract)

Session 3B: Sensing
Chairs: Armando Astarloa and Yolanda Lechuga
11:00 Andrea Sannino, José Ignacio Artigas and Aránzazu Otín
An Improved Discrete Time Amplifier-Less Potentiostat Architecture for Metabolic Sensing Applications (abstract)
11:30 Michel Justino Bai, Iñigo Adin Marcos and Markos Losada Gobantes
Learning to Sense Sustainably: RL-Based Control for Solar-Powered IoT Nodes (abstract)
12:00 Abel Reyes Cubas, David Galante Sempere and Javier Del Pino Suárez
A built-in CMOS temperature sensor for On-Chip Thermal Monitoring from 0ºC to 100ºC with a 0.137ºC of Innacuracy (abstract)

Session 4A: Security and Power Systems
Chairs: Jorge Portilla and Luis Parrilla
12:30 Sergio Tejeda Campos, Roberto Román, Rosario Arjona and Iluminada Baturone
Electromagnetic Side-Channel Attack on a Cloud-Based Fingerprint Recognition System (abstract)
13:00 Erica Tena-Sánchez, Francisco Eugenio Potestad Ordóñez, Miguel Martín-González, Alejandro Casado-Galán and Antonio J. Acosta
Low Entropy Masking Protection Scheme for ASCON Cipher to Counteract Side-Channel Attacks (abstract)
13:15 Carlos Fernández-García, Carlos J Jiménez-Fernández, Pilar Parra Fernández and Carmen Baena Oliva
A Lightweight AES Peripheral for RISC-V Cores and IoT Applications (abstract)
13:45 Pilar Cano-Lozano, Enrique Personal, Diego Francisco Larios, Samuel Dominguez-Cid, Juan Ignacio Guerrero and Carlos León
Electric vehicle emulator for study as a Distributed Energy Resource (abstract)

Session 4B: Signal Processing and Power Systems
Chairs: José Machado Da Silva and Luis Entrena
12:30 Santiago Murano, Martin Alejandro Colombo, Carlos De Marziani, Rubén Nieto Capuchino and Sofia Micaela Laskowski Orlandi
FPGA Architectures for Reliable Transmission of Pre-Stored Acoustic Signals in Underwater Localization Systems (abstract)
13:00 Fermin Esparza and Antonio Lopez-Martin
CMOS Micropower Current-Mode Sinh-Domain Filter with Multidecade Tuning (abstract)
13:30 Óliver P. Westin, Rafael T. Inuoe, Anderson A. Dionizio, Leonardo P. Sampaio and Sérgio A. Oliveira da Silva
Improved Modified Zeta Inverter for Single-Phase Grid-Tied System (abstract)

Session 5A: AI Circuits and Systems
Chairs: Juan Carlos López and Fernando Herrera
15:30 Marina Cordovilla Serrano, Pablo Sánchez Espeso and Andrés Martínez Lozano
Efficient Neural Architectures for Acoustic Monitoring of Livestock (abstract)
16:00 Rubén Padial-Allué, Alberto Martín-Martín, Encarnación Castillo, Uwe Meyer-Baese, Víctor Toral, Luis Parrilla and Antonio García
1-D Convolutional Autoencoder for Fetal and Maternal ECG Classification Oriented to Hardware Implementation Acceleration (abstract)
16:30 Jordi Fornt, Pau Fontova-Musté, Martí Caro, Jaume Abella, Josep Altet, Antonio Rubio and Francesc Moll
Approximate Circuits versus Quantization for Energy Efficient Deep Neural Networks (abstract)

Session 5B: Neuromorphic Circuits, Systems and Technologies I
Chair: Koldo Basterretxea and Antonio Rubio
15:30 Josep L. Rosselló, Christian F. Frasser, Arnau Salas-Barenys, Joan Cesari, Vincent Canals, Lluc Crespí-Castañer, Joan Font-Rosselló, Alejandro Morán Costoya and Miquel Roca
A 1.12 TOPS/W 180nm Stochastic Computing-based Neuromorphic Circuit (abstract)
16:00 Alejandro Morán, Lluc Lluc Crespí-Castañer, Christiam F. Franco, Joan Font-Rosselló, Vincent Canals, Miquel Roca and Josep L. Rosselló
A Comparative Analysis of Bipolar and Sign-Magnitude Stochastic Computing Approaches in Quantized Neural Networks (abstract)
16:30 Didac Llobet, Ioannis Chatzipaschalis, Antonio Calomarde and Antonio Rubio
Character Recognition Application of a Neural Circuit Including Lateral Inhibitory Mechanisms (abstract)

Session 6A: Neuromorphic Circuits, Systems and Technologies II
Chairs: Roberto Sarmiento and Hector Posadas
17:00 Prathamesh Satish Deshpande, Giovanni Grandi, Stephan Schoenfeldt and Fabian Lurz
Full-Integer Spiking Neural Network Inference with RISC-V ISA Extensions for Radar-based Gesture Recognition (abstract)
17:30 Bernabe Linares-Barranco, Luis Alejandro Camuñas Mesa and Teresa Serrano-Gotarredona
Three decades of IMSE Neuromorphic Engineering Group (abstract)
18:00 Cristina Bermúdez Martín, Samuel López Asunción and Pablo Ituero
Design Space Exploration of FPGA-Based Spiking Neural Networks for Angle of Arrival Detection (abstract)

Session 6B: System-Level Analysis and Exploration
Chair: Antonio Rubio and Koldo Basterretxea
17:00 Markel Galarraga, Charles-Alexis Lefebvre, Jon Perez-Cerrolaza and Jose A. Pascual
Analyzing Linux System Call Variability: Real-Time Patch Impact and System Call Monitoring (abstract)
17:30 Jose Luis Mira Serrano, Carlos Ernesto Hernandez Orellana, Jesús Barba Romero, María Soledad Escolar Diaz, Jose Antonio de la Torre Las Heras and Fernando Rincon Calle
Exploring Design Spaces in Embedded Systems: An Approach Based on Genetic Programming, Particle Swarm and Reinforcement Learning (abstract)
18:00 Eduardo Tomasi, César Fuguet, Christian Fabre and Frédéric Pétrot
HPC Workload Analysis Using Distributed Cross-ISA Binary Instrumentation (abstract)

Thursday, November 27th
Session 8 Keynote 2 : Alberto Bosio (Ecole Centrale de Lyon): Trustworthy AI: The role of the Hardware
Chair: Francisco Fernandez
Abstract: In the recent years, AI algorithms became so accurate that started to surpass human being in a wide range of tasks and are now currently used in applications that would be considered science-fiction only 10 years ago. However, everything as a cost and, for AI, it corresponds to the tremendous requirements in terms of computational power that driven the development of specialized hardware architectures (i.e., hardware accelerator) for DL workloads. One of the most popular hard- ware accelerator is the Systolic Array (SA) architecture, which is suitable to run inferences with low energy, low latency and high throughput. In particular, there exists virtually infinite types of SA in terms of dataflow depending on the kernel they have to run. If from one hand SA dataflow has been widely studied to improve energy efficiency and performance, the impact on the resilience to hardware faults has been neglected. Hardware faults can indeed jeopardize the execution of DL kernels leading to miss- classification and eventually to dramatic impacts when DL are used in safety-critical applications, such as autonomous driving. This talk will discusses the main consequences on the choice of a given hardware architecture to achieve by design Trustworthiness.

Session 9A: AI-Driven Development of High-Performance Electronic Systems and Applications-1
Chairs: Pablo Sanchez and Jesús Barba
11:00 Daniel Suárez, Pedro Hernández, Víctor Fernández and Gustavo Marrero
Video Action Recognition in SoC FPGAs driven by Neural Architectural Search (abstract)
11:30 Jaime Sancho, Manuel Villa, Gonzalo Rosa-Olmeda, Alejandro Martínez de Ternero, Miguel Chavarrías, Eduardo Juárez and César Sanz
Deep Learning-Based Depth Estimation for Facial Morphology Characterization in Neurosurgery Applications (abstract)
12:00 Félix David Suárez Bonilla, Gustavo Liñán-Cembrano and José Manuel De la Rosa Utrera
Multi-Domain Feature Extraction for ML-Based Over-the-Air RF Signal Classification (abstract)

Session 9B: Monitoring and Control
Chairs: María J. Avedillo and Francisco Fernandez
11:00 Rafaella Fiorelli, Juan Núñez Martínez and Eduardo Peralías
Nano-Oscillator Output Signal Monitoring Technique: Method and Device Implementation (abstract)
11:30 Amadeo de Gracia Herranz, Borja Gutierrez de Cabiedes, Javier de Mena Pacheco and Marisa Lopez-Vallejo
Ultra-Narrow Current Pulses Measurement Using a Cost-Effective Instrumentation System (abstract)
11:45 Juan Luis Soler-Fernandez, Angel Diéguez, Joan Daniel Prades and Oscar Alonso
Sub-nW Thyristor Based Wake-Up Timer for Low Duty Cycle IoT Sensing Applications (abstract)
12:15 Chih-Feng Wu, Bo-An Lin and Chi-Tien Sun
A 4×4 K-best Spatial Modulation MIMO Detection for Visible Light Communication Systems (abstract)

Session 10A: Circuit Design and Analysis
Chairs: Antonio Torralba and Miquel Roca
12:30 Juan Núñez, Rafaella Fiorelli and María J. Avedillo
Digital Ising-Based Solver for Scalable Max-Cut Optimization (abstract)
13:00 Muhammad Umer Khalid, Trond Ytterdal and Snorre Aunet
Comparative Analysis of Full Adders based on DTMOS Schmitt-Trigger Standard Cells Operating at Sub-100 mV Supply Voltage (abstract)
13:30 José Manuel Cruz Acosta, Irene Merino-Fernández, Javier del Pino and Sunil L. Khemchandani
Machine Learning-Based Physical Design of RFIC Transformers (abstract)

Session 10B: Powering circuits
Chairs: Alfio Dario Grasso and Aranzazu Otin
12:30 Mukul Agarwal, Nikhil Chourasiya, Sai Sumanth Pothuri and Subodh Prakash Taigor
Novel methodology for optimization of Charge Pump efficiency (abstract)
13:00 Mehdi Shahabi, Andoni Beriain Rodríguez and Noemi Perez Hernandez
A 65nm CMOS Ultra-Low-Quiescent-Current On-Chip PMIC for Energy-Limited Harvesting Systems (abstract)
13:30 Maria Clara Simões, Floriberto Lima, Marcelino Santos and Fábio Passos
A 300mA Fully-Integrated Inverter-Based LDO with Enhanced Supply Insensitivity for Smart Edge AI Applications (abstract)

Session 11: Industry-University Collaboration and Technology Transfer
In this panel, the MicroNanoSpain Competence center will be presented. Then, the panelists will presents their points of view about industry-university cooperation opportunities in electronic and microelectronic research and training:

  • Alfonso Gabarron, AESEMI & MicroNanoSpain Competence Center
  • Ana Peláez, Maxwell
  • Eduardo Casanueva, INDRA
  • Constantino Ruiz, AWGE

Chair: Antonio Lopez-Martin

Friday, November 28th
Session 13 Keynote: Ivan Rebollo (NXP): Dual Perspective on Non-Volatile Memory and RFID: Technological Progress and Engineering Challenges
Chair: Roc Berenguer
Abstract: Non-volatile memory (NVM) has played a pivotal role in the advancement of embedded systems, enabling persistent data storage across a wide spectrum of applications. The journey began with the development of electrically erasable programmable read-only memory (EEPROM) and multi-time programmable (MTP) technologies, which laid the foundation for reliable, low-power memory solutions in constrained environments. These early NVM architectures offered modest density and endurance but were instrumental in enabling secure data retention in analog and mixed-signal systems.
The emergence of Flash memory marked a significant leap in NVM capabilities, introducing higher density, faster access times, and scalable architectures suitable for both standalone and embedded applications. Flash variants became ubiquitous across consumer electronics, automotive systems, and industrial platforms. However, as process nodes scaled down and system requirements evolved, traditional Flash technologies began to encounter limitations in terms of write endurance, retention, and integration complexity—especially in ultra-low power and cost-sensitive domains.
This keynote provides a comprehensive overview of the current NVM landscape, with a particular emphasis on technologies derived from the MTP and Flash lineage applied to RFID (Radio-Frequency Identification) systems. We examine the trade-offs between endurance, retention, power consumption, and die area, and how these factors influence the selection of NVM in embedded designs.
In RFID (Radio-Frequency Identification) systems, where memory performance is tightly coupled with tag functionality, cost, and energy efficiency. RFID tags—especially passive and semi-passive types—demand memory solutions that operate reliably under stringent power and environmental constraints. EEPROM and MTP-based NVMs remain dominant in this space due to their proven reliability, low-voltage operation, and compatibility with analog front-end circuitry.
We analyze the specific memory requirements of RFID ICs, including write cycle limitations, retention across temperature extremes, and the need for secure data storage in authentication and anti-counterfeiting applications. The integration of NVM into RFID chips presents unique challenges, such as minimizing silicon footprint, ensuring process compatibility, and achieving cost targets for high-volume deployment.
The paper concludes with a discussion on future directions in NVM development for RFID and other embedded applications. Topics include the refinement of Flash and MTP architectures, and innovations in low-voltage programming techniques. As RFID continues to expand into new domains—from supply chain logistics to healthcare and smart packaging—the role of optimized NVM will be increasingly critical in enabling secure, efficient, and scalable solutions.

Session 14A: Hardware Accelerators
Chairs: Javier Uceda and Jaime Jiménez
11:00 Francisco Albertuz and Mario Garrido
Hardware-Efficient Gaussian and Sobel Filters for Real-Time Image Processing on FPGA (abstract)
11:30 Lluís Ribas-Xirgo
Hardware implementation of the Hungarian algorithm for optimum task assignments (abstract)
12:00 Pablo Hormigo-Jimenez and Javier Hormigo
Configurable Ultra-High-Throughput QRD FPGA Accelerators for small matrices (abstract)
12:15 Rubén Nieto, Silvia Iniesta, Santiago Murano, Pedro R. Fernández and Susana Borromeo
FPGA-Based Implementation of sEMG Feature Extraction and Movement Classification with MLP (abstract)

Session 14B: RF & Communications
Chairs: Roc Berenguer and Jose Ángel Miguel Díaz
11:00 Uxua Esteban-Eraso, Gesler Ramos, Santiago Celma, Francisco José Torcal-Milla and Carlos Sanchez-Azqueta
Design of a CMOS Transmitter Chain for Satellite on the Move Communications (abstract)
11:30 Alvaro Urain, David Del Rio, Andoni Beriain, Hector Solar, Roc Berenguer and Aleksei Nerushenko
Design of a 160-210 GHz SiGe HBT Square-Law Detector for Total Power Radiometers (abstract)
12:00 F. Bonfiglio-Buendía, Natalia-Abel Fernández-García, P. López, Victor M. Brea and Diego Cabello
CMOS SPDT Switch Topologies in the Frequency Range of 6 to 20 GHz (abstract)

Session 15A: AI-Driven Development of High-Performance Electronic Systems and Applications-2
Chairs: Pablo Sanchez and Soledad Escolar
12:30 Juan Gallego, José Ferreira, Luís Alves, Daniel Vázquez, João Bispo, Alfonso Rodríguez, Nuno Paulino and Andrés Otero
Acceleration of C/C++ Kernels and ONNX Models on CGRAs with MLIR-Based Compilation (abstract)
13:00 Maryam Katebzadeh, Daniel Vaquez, Andres Otero and Alfonso Rodriguez
A Framework for Automated CGRA Design Space Exploration with Genetic Algorithm Optimization (abstract)
13:30 Irene Merino-Fernandez, José Manuel Cruz Acosta, Javier del Pino and Sunil L. Khemchandani
Machine Learning for Microwave Pixelated Structures Design (abstract)

Session 15B: Quantum and low power
Chairs: Eduard Alarcon and Jose Maria Lopez Villegas
12:30 Aleksei Nerushenko, Hector Solar, Roc Berenguer and Alvaro Urain
A 1.15 mW SiGe BiCMOS Cryogenic LNA for Superconducting Qubit Readout with 4.5 K Noise Temperature from 4 to 9 GHz (abstract)
13:00 Ainhoa Leal, Luis Montal, Aleksei Nerushenko, Hector Solar and Roc Berenguer
A Methodology for Cryogenic Modeling of CMOS Technology Based on BSIM-BULK (abstract)
13:15 Muhammad Umer Khalid, Trond Ytterdal and Snorre Aunet
Robust DTMOS Schmitt-Trigger Circuits in 130 nm SOI CMOS for Sub-100 mV Supply Voltage (abstract)
13:45 Tom Bergmann, Joel Damiens, Alfonso Hildebrand Rueda, Stephane Lacouture, Remy Cellier and Nacer Abouchi
A Programmable, Negative, and Dynamically Biased Sampler for Ultra-Low Power Body-Bias Generators in 18nm FD-SOI (abstract)

Posters (during Coffee Break)

Juan Manuel Galan Serrano (CEIT-Basque Research and Technology Alliance (BRTA)), Javier Alonso Tirapu (CEIT-Basque Research and Technology Alliance (BRTA)), Andoni Irizar Picon (CEIT-Basque Research and Technology Alliance (BRTA)) and Ainhoa Cortes Vidal (CEIT-Basque Research and Technology Alliance (BRTA)).
VEC32-V: A versatile, extendible and configurable 32b RISC-V based MCU.
Abstract. Since the appearing of RISC-V ISA, a revolution has taken place in the open-hardware landscape, experiencing substantial growth in recent years. Numerous RISC-V cores have been developed to suit a wide range of use cases and applications. However, the number of fully functional microcontrollers based on these cores remains relatively limited. Furthermore, among the ones available, there exists a gap in configurability and target flexibility, which makes the process of tailoring the microcontroller to specific user requirements more complex. To address this gap, we present a configurable and flexible RISC-V microcontroller based on CV32E40P core from OpenHardware Foundation. It aims to facilitate easier customization and adaptation for a wide range of applications, ensuring that users can optimize performance and resources according to their specific needs.

Wael Elmanhawy (Product Management Director) and Pranit Oza (Product Engineer).
Automated SPICE Consolidation for Faster Circuit Verification.
Abstract. Hierarchical digital designs involve multiple levels of blocks and cells, leading to duplicated files that must be accessed repeatedly across shared networks. This duplication, exacerbated by network traffic fluctuations, significantly delays circuit verification runtimes. Manual approaches for consolidating SPICE netlists introduce risks like file size limits, naming collisions, and errors. This paper introduces an automated SPICE consolidation feature integrated into a Verilog-to-LVS (V2LVS) utility. [1] By automating netlist compilation and consolidation, design teams can reduce manual intervention, eliminate network duplication penalties, and improve verification performance. Real-world results demonstrate notable improvements in runtime efficiency, disk space usage, and memory requirements, representing a critical advancement for faster, more reliable circuit verification.

Sara Portero (ICCUB – University of Barcelona), David Mazzanti (ICCUB – University of Barcelona), Juan José Silva (ICCUB – University of Barcelona), Jose Maria Fernandez (ICCUB – University of Barcelona) and David Gascón (ICCUB – University of Barcelona).
Front End Readout Solutions for TOF-PET systems employing analog SiPMs.
Abstract. Timing resolution is of paramount importance in Positron Emission Tomography (PET) applications, and many other Time Of Flight systems such as LIDAR or mass spectrometry, since it directly translates to better image quality. This work will study what are the most relevant parameters (noise, bandwidth, parasitics, etc.) of a Front End readout circuit to optimize timing resolution. This work will include simulations and measurements.

José Angel Flores Bravo (UPV/EHU), Armando Astarloa (UPV/EHU), Jesús Lázaro (UPV/EHU), Aitzol Zuloaga (UPV/EHU) and Carlos Cuadrado (UPV/EHU).
Applications of Multipoint Distributed Fiber Optic Sensors for Monitoring in Critical Systems.
Abstract. Multipoint Distributed Fiber Optic Sensors (DFOS) have emerged as a transformative technology for monitoring critical systems across diverse industries, including aerospace, energy, environmental and infrastructure sectors. These sensors leverage advanced optical techniques, such as Rayleigh, Brillouin, Raman scattering, and interferometry, to provide high-resolution real-time data over vast distances. When integrated with edge computing systems, DFOS achieves enhanced real-time data processing capabilities, allowing efficient monitoring of multiple parameters such as temperature, strain, pressure, and vibration simultaneously and at discrete locations. This integration is particularly beneficial for critical applications that require rapid response times and robust performance in extreme environments. This paper presents a comparative analysis of multipoint DFOS architectures across diverse critical systems. Introduces a cross-sector technical evaluation based on key performance metrics, including spatial resolution, latency, and energy consumption. Provides actionable insights for matching sensor technologies with application-specific requirements in aerospace, civil infrastructure, industrial automation, environmental monitoring, transport, and chemical process control. This work proposes a novel integration framework that combines DFOS with edge computing architectures to improve responsiveness, scalability, and energy efficiency. This vision sets the foundation for next-generation innovative sensing systems that support real-time decision making in critical environments.

Kerman Pérez (UPV/EHU), Alejandro Arteaga (UPV/EHU), Jaime Jiménez (UPV/EHU), Armando Astarloa (UPV/EHU) and Jesús Lázaro (UPV/EHU).
Enriching Spacewire Virtual Testbeds with Python-based Simulation Tools.
Abstract. This research contributes to a Python-based simulation testbed for SpaceWire HDL IPs verification and standard conformance assessment. SpaceWire is a standard protocol for onboard spacecraft communications. Due to the criticality of this element, each SpaceWire implementation requires detailed and conscientious simulations that may vary depending on the components involved in the network. Thus, traditional HDL language simulation testbenches lack the flexibility to adapt quickly to different setups to offer productivity and ensure a reasonable time-to-market of the product. New technologies, like Python-based simulation tools, help to improve the verification stage. The simulation testbed presented in the paper explores this alternative, developing a fully operational setup in CocoTB and analyzing the simulation results to transition to HDL-based testbed implementations.

Hasnain Raza (University of Granada), Luis Parrilla (University of Granada), Encarnacion Castillo (University of Granada), Ruben Padial (UGR) and Antonio García (University of Granada).
Area-Efficient Point-Wise Multiplication for Crystals-Kyber.
Abstract. National Institute of Standard and Technologies (NIST) for Post-Quantum Cryptography (PQC). The proposed multiplier addresses the computational bottlenecks of polynomial multiplication, which is crucial for key generation, encryption, and decryption in Kyber. By improving the usual bit-reversal mechanism and integrating a Montgomery reduction method for modular reductions, the design significantly enhances both speed and area efficiency. Implemented on low-cost Field Programmable Gate Arrays (FPGAs), the multiplier achieves high performance with minimal hardware resources. The paper details the architecture, including the base case multiplier and modular reduction units, and compares the implementation results to existing methods. The proposed design demonstrates superior area-time product figures, making it suitable for real-world post-quantum cryptographic applications. This work contributes to the development of efficient hardware implementations of PQC algorithms, ensuring security against quantum computational threats.

Francisco Eugenio Potestad-Ordóñez (Escuela Politécnica Superior/Instituto de Microelectrónica de Sevilla (IMSE-CNM-CSIC-US)), Miguel Martín-González (Instituto de Microelectrónica de Sevilla (IMSE-CNM-CSIC-US)), Alejandro Casado-Galán (Instituto de Microelectrónica de Sevilla (IMSE-CNM-CSIC-US)) and Erica Tena-Sánchez (Escuela Politécnica Superior/Instituto de Microelectrónica de Sevilla (IMSE-CNM-CSIC-US)).
Error Detection and Correction Hardware Protection for ASCON cipher against fault analysis attacks.
Abstract. ASCON cipher was selected as the finalist in the NIST lightweight standardization call. Since the beginning of the standardization process, the cipher has been studied in detail, applying hardware and software fault injection attacks, compromising its security. This paper focuses on developing a countermeasure for hardware implementations related to error detection and correction for the ASCON cipher. In addition to detecting whether the device is under attack, the proposed scheme aims to correct possible detected faults. In the context of fault coverage and detection, extensive testing has been performed where the protection scheme consistently identifies faults introduced during permutations, achieving a 100% rate of fault detection and correction. This is achieved with a 69% increase in area overhead and negligible frequency degradation.

Matheus Minelli de Carvalho (IMT, Telecom Paris, LTCI, Institut Polytechnique de Paris), Benjamin Cheymol (CNRS, IN2P3, LPSC, Univ. Grenoble Alpes), Lirida Naviner (IMT, Telecom Paris, LTCI, Institut Polytechnique de Paris) and Rodrigo Possamai Bastos (CNRS, Grenoble INP, TIMA, Univ. Grenoble Alpes).
Near-sensor testing method for observing radiation-induced soft errors in modern sensors.
Abstract. Today, modern COTS sensors integrate interface circuits with analog and digital parts, enabling them to deliver sampled data directly to a processing system. Together, sensor and processing system are prone to manifest radiation-induced soft errors. The starting point of such soft errors is not readily evident and can vary between the different electronic parts, both from the COTS sensor and the processing system. This work proposes a method for observing radiation-induced soft errors as close to a sensor as possible, separating false positives related to soft errors in the processing system used to this end. Our method is general in the sense that there is no restriction towards a specific application or a sensor topology. We also showcase a case-study regarding 3-axis COTS accelerometers submitted to neutron radiation, with experimental results of soft errors occurring in different redundant axes.

Zeyuan Hui (Universitat Autònoma de Barcelona), Ivan Zamora (Universitat Autònoma de Barcelona), Arantxa Uranga (Universitat Autònoma de Barcelona) and Nuria Barniol (Universitat Autònoma de Barcelona).
Design of an Element-Level Delay Line with Duty-Cycle Correction for a large PMUT Array.
Abstract. In modern ultrasound imaging systems, uniform excitation across a large-aperture piezoelectric micromachined ultrasonic transducer (PMUT) array is critical for high-resolution beamforming and image quality. This work presents the design of a 64-channel element-level delay architecture based on a voltage-controlled delay line (VCDL) integrated with a duty-cycle corrector (DCC) circuit. The proposed system enables precise per-channel delay tuning with a resolution of 5 ns and supports a maximum element delay of 19.80 ns for a 20MHz system. The DCC module effectively mitigates duty-cycle distortion, maintaining a coefficient of variation below 0.60% across all channels. Simulation results verify the feasibility and performance of the architecture, providing a promising solution for high-density PMUT-CMOS integration with reliable transmission-side beamforming.

Álvaro Falcón Santana (Universidad de Las Palmas de Gran Canaria), Carlos Vega García (Universidad de Las Palmas de Gran Canaria), Antonio José Rodríguez Almeida (Universidad de Las Palmas de Gran Canaria), Himar Fabelo Gómez (Universidad de Las Palmas de Gran Canaria), Pedro Francisco Pérez Carballo (Universidad de Las Palmas de Gran Canaria) and Gustavo Iván Marrero Callicó (Universidad de Las Palmas de Gran Canaria).
MRI segmentation and mesh generation targeting real-time brain-shift compensation in neurosurgery.
Abstract. During brain tumour surgeries, neurosurgeons face several challenges when addressing the intraoperative resection of the tumour. One of the most challenging problems is the brain-shift phenomenon, which is the deformation of the brain occurred after performing the craniotomy and durotomy, as well as during tumour removal. This phenomenon leads to a reduction in the accuracy of neuronavigation systems that rely on preoperatively acquired imaging data. This work introduces the first steps toward addressing this issue, targeting the generation of a biomechanical model capable of estimating the intraoperative brain deformations in real-time. We propose a method to generate a finite element model of the brain from a preoperative magnetic resonance image by using a pipeline that segments the different anatomical regions of the brain, converts them to a mesh model, and the definition of a finite element model based in the morphology and tissue properties. This work focuses on the generation of the anatomical and structural foundations required for the posterior brain shift simulations. In addition, since these models are computationally demanding, a profiling of the different stages of the pipeline was performed, identifying its potential bottlenecks to achieve real-time performance, and also suggesting feasible solutions for its hardware acceleration.

Lydia Bush-Espinosa (IC Málaga), Michael Gater (University of Nottingham), Carol de Benito (Universitat Illes balears), Stavros Stavrinides (Democritus University of Thrace), Miquel Roca (University of Balearic Islands), Salvador Dueñas (UNIVERSIDAD DE VALLADOLID), Álvaro Pineda (IC Málaga) and Rodrigo Picos (Universitat de les Illes Balears).
A programming system using current pulses for organic memristors deposited on the same chip.
Abstract. This paper presents the design and simulation of an integrated readout system in 0.18 µm CMOS technology for characterizing organic memristors using current. The memristive devices are to be fabricated post-process atop the chip’s passivation layer. The system comprises three key blocks: (1) a programmable current source with four selectable mirrors (nA to mA range), calibrated via external resistors and controlled by a 0–1 V input; (2) a memristor selector enabling bidirectional current flow through two memristor devices (MEM1/MEM2) with low-resistance switches (<1 $\Omega$); and (3) a logic block for device/polarity selection. Simulations demonstrate robust performance across process corners, with a maximum current of 3 mA (5 V supply), settling times <0.5 $\micro$ s for pulsed operation, and compatibility with organic memristors (1 k$\Omega$–1 M$\Omega$ range). The design addresses critical challenges in memristor integration on silicon, offering a scalable platform for hybrid CMOS-memristor systems in flexible electronics.

Amalia Arias (IMB-CNM (CSIC)), Manuel Lozano (IMB-CNM (CSIC)), Esteve Amat (IMB-CNM (CSIC)), Alvaro Pineda (Integrated Circuits Malaga S.L.), Bartomeu Servera (Integrated Circuits Malaga S.L.) and Angel Barbancho (Integrated Circuits Malaga S.L.).
Floating-Gate Based Hybrid Sensor for Low-Power CMOS-Compatible Radiation Dosimetry.
Abstract. This work presents the design of a passive floating-gate MOS (FGMOS) dosimeter sensor, developed within a hybrid architecture that separates the sensing and readout functions. The sensing structure is implemented using self-aligned transistors over LOCOS isolation, with a shared floating gate connected to a PMOS injector and an NMOS readout transistor. A reference cell with identical geometry and fixed gate potential is also included. The complete layout has been translated into a fabrication-ready mask set, fully compatible with standard CMOS technology. Fabrication will take place at the CMOS line of the IMB-CNM (CSIC). This development is the result of a close collaboration between academic and industrial institutions, combining expertise in microelectronic sensor design and integrated systems. A dedicated readout SoC developed by IC Málaga provides analog signal conditioning, recharge control, and digital processing, enabling accurate interpretation of the sensor response. This hybrid solution allows independent optimization of the sensor and readout paths, offering enhanced performance and flexibility compared to fully monolithic implementations.

Alfonso Guerrero-de-Mier (University of Seville), Pedro Blanco-Carmona (University of Seville), Ramón González-Carvajal (University of Seville) and Alejandro Rodríguez-González (Guadalquivir Hydrographic Confederation).
A Methodological Framework for Developing Digital Twins in River Basins.
Abstract. The purpose of this paper is to define and implement a comprehensive workflow for the development of digital twins of river basins. The proposed methodology was validated using the Guadalquivir River Basin in Spain as a case study, incorporating hydrological and meteorological data acquired automatically and manually. Various scenarios were tested to evaluate the workflow’s ability to handle different input conditions and feature sets. The results of this study demonstrate the potential of the proposed approach to facilitate the integration of digital twins into large-scale water resource management applications and optimize services.

Emili Manzano-Aguilar (University of the Balearic Islands), Bartomeu Alorda-Ladaria (University of the Balearic Islands), Carola De Benito (University of the Balearic Islands), Francisco J. Cereceda (University of the Balearic Islands) and Jesus Molina (University of the Balearic Islands).
Non-Invasive Data Extraction OCR based method for Commercial Medical Devices.
Abstract. Commercial medical devices present significant challenges for clinical research and protocol improvement due to their limited data sharing capabilities. This paper presents a noninvasive method based on Optical Character Recognition (OCR) to extract real-time data from Automated External Defibrillators (AEDs) implemented on accessible hardware platforms like the Raspberry Pi 5 and PiCamera2 mounted on a custom 3D-printed arm, precisely positioned using LiDAR modeling. The software implements image processing and OCR algorithms to capture, interpret, and transmit critical parameters such as EtCO2, SpO2, heart rate, and shock count. The proposed system demonstrates reliable performance with second-level data registration and high adaptability to various medical and non-medical applications. Furthermore, its seamless integration into the Internet of Things (IoT) ecosystems highlights its potential for broader use cases. This work provides a cost-effective, yet sophisticated approach to extracting valuable information from closed systems without compromising their functionality.

Joan Teruel (Barcelona Supercomputing Center / Universitat Politecnica de Catalunya), Elias Perdomo (Barcelona Supercomputing Center / Universitat Politecnica de Catalunya), David Castells (Barcelona Supercomputing Center / Universitat Autònoma de Barcelona), Xavier Martorell (Barcelona Supercomputing Center / Universitat Politecnica de Catalunya) and Behzad Salami (Barcelona Supercomputing Center).
Telemetry Tools for Alveo Accelerator Cards.
Abstract. Field Programmable Gate Arrays (FPGAs) are increasingly utilised as computational platforms in data centres and high-performance computing environments due to their reconfigurability and efficiency. This thesis explores the development of telemetry tools for AMD/Xilinx Alveo accelerator cards, showcasing the growing need for effective monitoring systems to characterise FPGA behaviour and profile FPGA-based designs. Telemetry data, such as temperature, power, and voltage, provides crucial insights into performance optimisation and energy efficiency. This bachelor’s thesis demonstrates the shortcomings of existing Out-of-Band (OOB) telemetry systems. Particularly, their incompatibility with modern infrastructures and incomplete documentation. As a solution, this work develops and implements an In-Band (IB) telemetry methodology, successfully integrating telemetry capabilities into the FPGA infrastructure used at the BSC. This work also introduces a novel approach for integrating telemetry into custom FPGA designs, enabling enhanced profiling and characterisation of specific modules. Showcasing the potential for reduced energy consumption, more efficient resource management, improved cooling strategies, and improved design efficiency in FPGA-based systems.