This paper discusses the introduction of an integrated Posit Processing Unit (PPU) as an alternative to Floating-point Processing Unit (FPU) for Deep Neural Networks (DNNs) in automotive applications. Autonomous Driving tasks are increasingly depending on DNNs. For example, the detection of obstacles by means of object classification needs to be performed in real-time without involving remote computing. To speed up the inference phase of DNNs the CPUs on-board the vehicle should be equipped with co-processors, such as GPUs, which embed specific optimization for DNN tasks. In this work, we review an alternative arithmetic that could be used within the co-processor. We argue that a new representation for floating point numbers called Posit is particularly advantageous, allowing for a better trade-off between computation accuracy and implementation complexity. We conclude that implementing a PPU within the co-processor is a promising way to speed up the DNN inference phase.


    Access

    Download


    Export, share and cite



    Headlight Range Estimation for Autonomous Driving using Deep Neural Networks

    Mayr, Jakob / Giracoglu, Can / Unger, Christian et al. | IEEE | 2019


    PATCHING DEPLOYED DEEP NEURAL NETWORKS FOR AUTONOMOUS MACHINE APPLICATIONS

    BOJARSKI MARIUSZ / MUELLER URS DR / BEAT FLEPP et al. | European Patent Office | 2022

    Free access

    An Autonomous Driving Approach Based on Trajectory Learning Using Deep Neural Networks

    Wang, Dan / Wang, Canye / Wang, Yulong et al. | Springer Verlag | 2021


    Exploiting Hierarchy for Scalable Decision Making in Autonomous Driving

    Sonu, Ekhlas / Sunberg, Zachary / Kochenderfer, Mykel J. | IEEE | 2018


    Distance Map Estimation of Stereoscopic Images Using Deep Neural Networks for Autonomous Vehicle Driving

    Taimouri, Vahid / Cordonnier, Michel / Lee, Kyoung Min et al. | British Library Conference Proceedings | 2017