Estimating the reliability of DNNs in the face of permanent GPU hardware failures Mr. Juan Balaguera, PhD Candidate, DAUIN Graphic Processing Units (GPUs) are crucial for modern Deep Neural Network (DNN) acceleration. However, these devices can be affectedby faults that might jeopardize the DNNs’ realiability. My researchproposes fault simulation strategies to effectively assess the impact of permanent defects on GPUs regardless of the softwareimplementation of the neural networks. Functional Stimuli Generation for Burn-In Test Mr. Nick Deligiannis, PhD Candidate, DAUIN In high-reliability applications, Burn-In testing (BI) is crucial to combat early failures. Traditional static BI is inefficient for modern dense circuits. In this work, we propose automated methods able to generate effective, functional stress inducing stimuli, especially for pipelined processors, destined for dynamic BI test. Reliability and Performance Challenges of Next-Generation Smart Power Battery Management Systems for Electric Mobility Mr. Amirhossein Ahmadi, PhD Candidate, DET This project discusses the impact of electromagnetic interference (EMI) on the battery management systems (BMS) and BMS vertical interface (VIF). The susceptibility to EMI is tackeld by transistor-level simulations and tests for the first time aiming to highlight the failure mechanisms and consequently to propose methods to enhance the performance. Room: Maxwell Room, Bldg: DET, Politecnico di Torino, Corso Duca Degli Abruzzi 24, Torino, Piemonte, Italy, 10129
This event has passed.
