Francesco Pessia
Papers
1
Total Citations
2
H-Index
1
About
Francesco Pessia is a researcher at the forefront of reliable and efficient computing, with a primary focus on the intersection of hardware reliability, machine learning, and high-performance architectures. His work critically examines how scheduling policies and resource management affect the dependability of Graphics Processing Units (GPUs) when executing complex workloads, particularly Convolutional Neural Networks (CNNs). In his highly cited 2024 study, Pessia systematically analyzed the impact of different scheduling strategies on GPU reliability during CNN operations, revealing how programming flexibility and parallelism—key advantages of GPUs—can introduce vulnerabilities in safety-critical applications. This contribution is vital as CNNs are increasingly deployed in autonomous systems and other domains where failure is not an option. By bridging the gap between performance optimization and fault tolerance, Pessia’s research provides essential guidelines for designing more robust accelerators. His work has already garnered attention, accumulating citations that underscore its relevance to both the architecture and machine learning communities. For students and researchers exploring dependable AI hardware, Francesco Pessia offers a compelling perspective on how to balance speed with safety in next-generation computing systems.
Research Focus
Key Achievements
Top Papers
- 1