Fernando Fernandes dos Santos

I’m a tenured researcher at Inria Rennes in the Taran Team. My main topics of research are fault tolerance and reliability aspects of RISC-V, GPUs, processing in memory, and AI accelerators. I earned a bachelor’s degree in Computer Science from Western Paraná State University, followed by master’s and PhD degrees from the Federal University of Rio Grande do Sul, Brazil. After completing my PhD, I joined Inria’s TARAN team in Rennes, France, as a postdoctoral researcher in the MSCA-BIENVENÜE program.

Publications

For a complete and up-to-date list of my publications, see my Google Scholar profile or my HAL CV.

Supervision

PhD students

• Lucas Roquet (2024–present): Dependability of large machine-learning models on GPUs
• Mario Wagner (2025–present): Reliability of emerging machine-learning accelerators
• Ioannis Zacharioudakis (2026–present): Reliability and protection strategies for neuromorphic accelerators

Master’s students

• Maxime Zingraff (2025–2026): Mitigating aging effects on Processing Using DRAM architectures
• Maxwell Pirtle (2024–2026): Time predictability and Worst-Case Execution Time analysis on NVIDIA GPUs
• Mathys Minsac (2025): Fault simulation and mitigation on Processing Using DRAM architectures
• El Mehdi Bel Haddad (2024): Reliability assessment of UPMEM Processing Near Memory accelerators
• Lucas Roquet (2022–2023): Fault simulation, reliability, and security of Vision Transformers

Undergraduate students

• Senakpon V Jivana F Gnansounou (2026): Reliability of compressed DNNs for microcontrollers
• Remi Loison (2025): Fault simulation on RISC-V processors
• Raphael Dennis (2024): Microarchitectural fault simulation for a commercial RISC-V processor
• Leah (2024): Fault simulation on CUDA-based applications
• Melanie Romano (2023): Reliability assessment of applications running on the GreenWaves GAP9 RISC-V processor
• Geronimo Veit Acosta (2021): Fault tolerance using reduced-precision duplication with comparison

Do you want to join?

I am open to supervising new students who are motivated, eager to learn, and interested in computer architecture. If you fit this description, please feel free to contact me via email. The topics I am willing to explore include, but are not limited to, the following:

• Reliability of parallel architectures: permanent, transient, and intermittent faults
• Radiation-induced effects on computers: soft error rates, accumulated dose, in-orbit computing, etc.
• Real-time applications: real-time GPUs, real-time constraints
• Processing in memory: reliability and optimization

Projects

• Reliable AI for the Next Generation of Safety- and Mission-Critical Embedded Systems (RAISE)
  • My role: Principal investigator
• Transformers’ Reliability for Safety-Critical Applications on Graphic Processing Units (TCHE)
  • My role: Principal investigator
• ARCHitectures based on unconventional accelerators for dependable/energY efficienT AI Systems (ARCHYTAS)
  • My role: Participating in the project
• REliable hardware for TRUSTworthy artificial INtelliGence (RE-TRUSTING)
  • My role: Participating in the project
• FAult-aware timing behaviour for safety-critical multicore SYstems (FASY)
  • My role: Participating in the project

Teaching

2025

• Introduction to C language and Unix (University of Rennes)
• TinyML Accelerators in conjunction with Olivier S., Angeliki K., Marcello T., and Romain F.
  • The materials of this class can be found at: TinyML GitLab
• Frugal AI Master Course (University of Rennes) in conjunction with Angeliki K. and Marcello T.
  • The materials of this course can be found at: Frugal AI GitLab

2024

• Introduction to C language and Unix (University of Rennes)
• TinyML Accelerators in conjunction with Olivier S., Angeliki K., Silviu F., Marcello T., and El-mehdi L.
  • The materials of this class can be found at: TinyML GitLab

2020–2021

Introduction to programming: C/Python for students in STEM courses at the Federal University of Rio Grande do Sul, Brazil.

Awards

2026

• DSN’26 Distinguished Reviewer Award

2023

• I am the semi-finals winner of the McCluskey Doctoral Thesis Award for my thesis titled Understanding and Improving GPUs’ Reliability Combining Beam Experiments with Fault Simulation at the IEEE European Test Symposium 2023.

• I won the McCluskey Doctoral Thesis Award at the 2023 IEEE International Test Conference (ITC).

2022

• Best 2021 Thesis on Computer Science of the whole Brazil by Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) Award. CAPES is the Brazilian government agency for research funding.

2021

• Paul Phelps Continuing Education Grant from the IEEE Nuclear and Plasma Sciences Society community, in the 2021 IEEE NSREC.

• My thesis was evaluated Cum Laude by a board of reviewers composed of Dr. Timothy Tsai (NVIDIA), Dr. Dimitris Gizopoulos (National and Kapodistrian University of Athens), and Dr. Evgenia Smirni (College of William & Mary). The text of the thesis is available (English only) at this link.

2018

• Best paper runner-up: In the 2018 IEEE/IFIP DSN, the paper Code-Dependent and Architecture-Dependent Reliability Behaviors co-authored by me was the runner-up for the best paper award.
• Best of SELSE: In the 2018 edition of SELSE, the paper Evaluation and Mitigation of Soft-Errors in Neural Network-Based Object Detection in Three GPU Architectures co-authored by me won the best paper award.

2014

• In my Bachelor’s in Computer Science, I graduated with honors. The Brazilian Computer Society also recognized me as the outstanding student of the year at UNIOESTE.