Juan David Guerrero Balaguera
Researcher | AI & ML | Parallel architectures | AI accelerators | Hardware designer | Fault tolerance | Design for Testability | Reliability | GPUs | FPGAs |
juan.guerrero at polito dot it
Hey! thanks for visiting my webpage
I am a Postdoctoral Research Fellow of the CAD group in the Department of Control and Computer Engineering DAUIN from Politecnico di Torino - PoliTo in Turin, Italy.
I completed my Ph.D (2024) in Computer and Control Engineering from PoliTo, where I was advised by Prof. Matteo Sonza Reorda and Prof. Ernesto Sanchez. During my Ph.D I worked on the reliability enhancement of GPU architectures. Also, I obtained twice the Ph.D quality awards of my cycle. This distinction is given by the doctoral school of PoliTo to the best Ph.D students every year. I obtained bachelor’s and master’s degrees in electronics engineering from Universidad Pedagogica y Tecnologica de Colombia in 2013 and 2017, respectively.
From 2014 to 2020, I was an assistant professor at the Electronics Engineering School of the Universidad Pedagogica y Tecnologica de Colombia, where I used to teach digital design, embedded systems, computer architecture, and image processing with FPGAs.
My research work focuses on the design of dependable hardware for safety-critical applications, comprising a wide spectrum of flavors, including but not limited to:
- Advanced digital design for FPGAs
- Computational arithmetic for AI
- Accelerators for vision applications
- Functional in-field testing of hardware components (e.g., CPUs, and GPUs)
- Dependability and fault tolerance for AI accelerators
- AI reliability assessment (i.e., error modeling of hardware faults).
selected publications
- ElectronicsExploring hardware fault impacts on different real number representations of the structural resilience of tcus in gpusElectronics, 2024
- SC’23Understanding the Effects of Permanent Faults in GPU’s Parallelism Management and Control UnitsIn Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis, 2023
- IEEE Design & TestSTLs for GPUs: using high-level language approachesIEEE Design & Test, 2023
- DATE’22A compaction method for STLs for GPU in-field testIn 2022 Design, Automation & Test in Europe Conference & Exhibition (DATE), 2022
- ITC’22A multi-level approach to evaluate the impact of GPU permanent faults on CNN’s reliabilityIn 2022 IEEE International Test Conference (ITC), 2022
- BEC’20Design and Verification of an open-source SFU model for GPGPUsIn 2020 17th Biennial Baltic Electronics Conference (BEC), 2020
- DynaFPGA-based translation system from colombian sign language to textDyna, 2015