Abstract
Software-defined, AI-enabled Autonomous Vehicles (AV) require Graphics Processing Units (GPUs) for developing and running AI and XR end-to-end. GPUs enable the technology to apply deep learning algorithms to complex sensor data in order to produce autonomous actions. In addition, GPUs are used for the creation of next-generation augmented reality (AR) interfaces, with information pertaining to points of interest, alerts, navigation, and co-pilot support for drivers. This webinar will provide an overview of the fundamental characteristics of GPUs required to support new functionalities, including architecture, design, and verification points of view. It will also describe how dependability and functional safety need to be considered in order to provide the required level of trust for GPU computing.
About the speaker
Riccardo Mariani is widely recognized as an expert in functional safety and integrated circuit reliability. In his current role as VP of Industry Safety at NVIDIA, he is responsible for driving safety alignment across NVIDIA’s automotive and embedded business units as industrial, robotics, and healthcare. To this end, he is responsible for developing cohesive safety strategies and cross-segment safety processes, architecture, and products that can be leveraged across NVIDIA’s AI-based hardware and software platforms. Mariani is 2020 First VP of IEEE Computer Society, re-elected for 2021 as well. He is also VP for Standardisation Activities of the same Society. He chairs the IEEE Special Technical Community on Reliable, Safe, Secure, and Time Deterministic Intelligent Systems as also the IEEE P2851 standard on safety analysis and verification. He spent the bulk of his career as CTO of Yogitech, an industry leader in functional safety technologies. Before co-founding the Italian company in 2000, he was technical director at Aurelia Microelettronica, where his responsibilities included leading high-reliability topics in projects with CERN in Geneva.