This work focuses on the time-predictable execution of Deep Neural Networks (DNNs) accelerated on FPGA System-on-Chips (SoCs). The modern DPU accelerator by Xilinx is considered. An extensive profiling campaign targeting the Zynq Ultrascale+ platform has been performed to study the execution behavior of the DPU when accelerating a set of state-of-the-art DNNs for Advanced Driver Assistance Systems (ADAS). Based on the profiling, an execution model is proposed and then used to derive a response-time analysis. A custom FPGA module named DICTAT is also proposed to improve the predictability of the acceleration of DNNs and tighten the analytical bounds. A rich set of experimental results based on both analytical bounds and measurements from the target platform is finally presented to assess the effectiveness and the performance of the proposed approach on ADAS applications.

Time-Predictable Acceleration of Deep Neural Networks on FPGA SoC Platforms

Restuccia F.;Biondi A.
2021-01-01

Abstract

This work focuses on the time-predictable execution of Deep Neural Networks (DNNs) accelerated on FPGA System-on-Chips (SoCs). The modern DPU accelerator by Xilinx is considered. An extensive profiling campaign targeting the Zynq Ultrascale+ platform has been performed to study the execution behavior of the DPU when accelerating a set of state-of-the-art DNNs for Advanced Driver Assistance Systems (ADAS). Based on the profiling, an execution model is proposed and then used to derive a response-time analysis. A custom FPGA module named DICTAT is also proposed to improve the predictability of the acceleration of DNNs and tighten the analytical bounds. A rich set of experimental results based on both analytical bounds and measurements from the target platform is finally presented to assess the effectiveness and the performance of the proposed approach on ADAS applications.
2021
Inglese
Proceedings - Real-Time Systems Symposium
2021-
441
454
14
978-1-6654-2802-6
Institute of Electrical and Electronics Engineers Inc.
New York
STATI UNITI D'AMERICA
42nd IEEE Real-Time Systems Symposium, RTSS 2021
deu
2021
Internazionale
no
273
2
info:eu-repo/semantics/conferenceObject
4 Contributo in Atti Convegno (Proceeding)::4.1 Contributo Atti Congressi/Articoli in extenso
open
Restuccia, F.; Biondi, A.
File in questo prodotto:
File Dimensione Formato  
RTSS21.pdf

accesso aperto

Tipologia: Documento in Pre-print/Submitted manuscript
Licenza: Copyright dell'editore
Dimensione 355.97 kB
Formato Adobe PDF
355.97 kB Adobe PDF Visualizza/Apri

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11382/545030
 Attenzione

Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte dell'ateneo

Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 17
social impact